Hypergraph neural networks (HGNNs) have shown great potential in modeling higher-order relationships among multiple entities. However, most existing HGNNs primarily emphasize low-pass filtering while neglecting the role of high-frequency information. In this work, we present a theoretical investigation into the spectral behavior of HGNNs and prove that combining both low-pass and high-pass components leads to more expressive and effective models. Notably, our analysis highlights that high-pass signals play a crucial role in capturing local discriminative structures within hypergraphs. Guided by these insights, we propose a novel sheaflet-based HNNs that integrates cellular sheaf theory and framelet transforms to preserve higher-order dependencies while enabling multi-scale spectral decomposition. This framework explicitly emphasizes high-pass components, aligning with our theoretical findings. Extensive experiments on benchmark datasets demonstrate the superiority of our approach over existing methods, validating the importance of high-frequency information in hypergraph learning.

Hypergraphs provide a natural and expressive framework for modeling high-order relationships, enabling the representation of group-wise interactions beyond pairwise connections. While hypergraph neural networks (HNNs) have shown promise for learning on such structures, existing models often rely on shallow message passing and lack the ability to extract multiscale patterns. Framelet-based techniques offer a principled solution by decomposing signals into multiple frequency bands. However, most prior framelet systems, particularly Haar-type ones, are sensitive to node ordering and fail to ensure consistent representations under permutation, leading to instability in hypergraph learning. To address this, we propose Permutation Equivariant Framelet-based Hypergraph Neural Networks (PEF-HNN), a novel framework that integrates multiscale framelet analysis with permutation-consistent learning. We construct a new family of permutation equivariant Haar-type framelets specifically designed for hypergraphs, supported by theoretical analysis of their stability and decomposition properties. Built upon these framelets, PEF-HNN incorporates both low-pass and high-pass components across multiple scales into a unified neural architecture. Extensive experiments on nine benchmark datasets, including three homophilic and four heterophilic hypergraphs, as well as two real-world datasets for visual object classification, demonstrate the effectiveness of our approach, consistently outperforming existing HNN baselines and highlighting the advantages of permutation equivariant framelet design in hypergraph representation learning.

Time series forecasting requires architectures that simultaneously achieve three competing objectives: (1) strict temporal causality for reliable predictions, (2) sub-quadratic complexity for practical scalability, and (3) multi-scale pattern recognition for accurate long-horizon forecasting. We introduce AutoHFormer, a hierarchical autoregressive transformer that addresses these challenges through three key innovations: 1) Hierarchical Temporal Modeling: Our architecture decomposes predictions into segment-level blocks processed in parallel, followed by intra-segment sequential refinement. This dual-scale approach maintains temporal coherence while enabling efficient computation. 2) Dynamic Windowed Attention: The attention mechanism employs learnable causal windows with exponential decay, reducing complexity while preserving precise temporal relationships. This design avoids both the anti-causal violations of standard transformers and the sequential bottlenecks of RNN hybrids. 3) Adaptive Temporal Encoding: a novel position encoding system is adopted to capture time patterns at multiple scales. It combines fixed oscillating patterns for short-term variations with learnable decay rates for long-term trends. Comprehensive experiments demonstrate that AutoHFormer 10.76X faster training and 6.06X memory reduction compared to PatchTST on PEMS08, while maintaining consistent accuracy across 96-720 step horizons in most of cases. These breakthroughs establish new benchmarks for efficient and precise time series modeling. Implementations of our method and all baselines in hierarchical autoregressive mechanism are available at https://github.com/lizzyhku/Autotime.

The significant features identified in a representative subset of the dataset during the learning process of an artificial intelligence model are referred to as a 'global' explanation. Three-dimensional (3D) global explanations are crucial in neuroimaging, where a complex representational space demands more than basic two-dimensional interpretations. However, current studies in the literature often lack the accuracy, comprehensibility, and 3D global explanations needed in neuroimaging and beyond. To address this gap, we developed an explainable artificial intelligence (XAI) 3D-Framework capable of providing accurate, low-complexity global explanations. We evaluated the framework using various 3D deep learning models trained on a well-annotated cohort of 596 structural MRIs. The binary classification task focused on detecting the presence or absence of the paracingulate sulcus (PCS), a highly variable brain structure associated with psychosis. Our framework integrates statistical features (Shape) and XAI methods (GradCam and SHAP) with dimensionality reduction, ensuring that explanations reflect both model learning and cohort-specific variability. By combining Shape, GradCam, and SHAP, our framework reduces inter-method variability, enhancing the faithfulness and reliability of global explanations. These robust explanations facilitated the identification of critical sub-regions, including the posterior temporal and internal parietal regions, as well as the cingulate region and thalamus, suggesting potential genetic or developmental influences. For the first time, this XAI 3D-Framework leverages global explanations to uncover the broader developmental context of specific cortical features. This approach advances the fields of deep learning and neuroscience by offering insights into normative brain development and atypical trajectories linked to mental illness, paving the way for more reliable and interpretable AI applications in neuroimaging.

Healthcare decision-making requires not only accurate predictions but also insights into how factors influence patient outcomes. While traditional machine learning (ML) models excel at predicting outcomes, such as identifying high-risk patients, they are limited in addressing "what if" questions about interventions. This study introduces the Probabilistic Causal Fusion (PCF) framework, which integrates Causal Bayesian Networks (CBNs) and Probability Trees (PTrees) to extend beyond predictions. PCF leverages causal relationships from CBNs to structure PTrees, enabling both the quantification of factor impacts and the simulation of hypothetical interventions. The framework is evaluated on three clinically diverse, real-world datasets, MIMIC-IV, Framingham Heart Study, and BRFSS (Diabetes), demonstrating consistent predictive performance comparable to conventional ML models, while offering enhanced interpretability and causal reasoning capabilities. In contrast to conventional approaches focused solely on prediction, PCF offers a unified framework for prediction, intervention modelling, and counterfactual analysis, forming a holistic toolkit for clinical decision support. To enhance interpretability, PCF incorporates sensitivity analysis and SHapley Additive exPlanations (SHAP). Sensitivity analysis quantifies the influence of causal parameters on outcomes such as Length of Stay (LOS), Coronary Heart Disease (CHD), and Diabetes, while SHAP highlights the importance of individual features in predictive modelling. This dual-layered interpretability offers both macro-level insights into causal pathways and micro-level explanations for individual predictions. By combining causal reasoning with predictive modelling, PCF bridges the gap between clinical intuition and data-driven insights. Its ability to uncover relationships between modifiable factors and simulate hypothetical scenarios provides clinicians with a clearer understanding of causal pathways. This approach supports more informed, evidence-based decision-making, offering a robust framework for addressing complex questions in diverse healthcare settings.

MOTIVATIONS: Electroencephalography (EEG) is a non-invasive method that records brain electrical activity from scalp electrodes, offering millisecond temporal resolution but limited spatial detail due to sparse sensor layouts. RESULTS: We present DiBiMa-EEGSR, a bidirectional Mamba-2 diffusion framework for spatio-temporal EEG super-resolution that reconstructs high-resolution signals from standard low-density recordings without additional hardware. The method formulates super-resolution as conditional generative inference and integrates a diffusion process with a bidirectional state-space backbone to model long-range temporal dependencies with linear complexity. Conditioning on low-resolution inputs, electrode positions and task labels enables anatomically coherent and context-aware reconstruction. A one-step sampling strategy substantially reduces inference time while preserving fidelity. Across two public benchmarks, the approach improves reconstruction accuracy, spatial coherence and spectral preservation over convolutional, transformer-based and prior diffusion models in both spatial and temporal upsampling tasks, providing a scalable pathway toward high-resolution electrophysiological imaging. AVAILABILITY AND IMPLEMENTATION: Code to reproduce ablation experiments, training and evaluation of the proposed BiMa and DiBiMa EEGSR models are available at https://github.com/UgoLomoio/DiBiMa-EEGSR.git. Model weights are available at https://huggingface.co/Ugo96/DiBiMa-EEGSR while an interactive demo for EEG spatial super-resolution using our models can be found at https://huggingface.co/spaces/Ugo96/DiBiMa-EEGSR-Demo.

Ensuring robust cybersecurity in modern network environments is increasingly challenging due to the growing complexity and volume of network traffic data. Traditional detection systems often fail to identify stealthy and sophisticated attacks, such as Distributed Denial of Service (DDoS), ARP poisoning, and reconnaissance scans. Moreover, many existing methods lack transparency and produce reports that are difficult for analysts to interpret, slowing both threat comprehension and response. This paper addresses these challenges by introducing a novel methodology that integrates Knowledge Graphs, XAI techniques and Large Language Models (LLMs) to enhance network threat detection, classification, explainability, and automated reporting. The proposed approach employs Graph-BERT to encode complex communication patterns and semantic relationships into enriched knowledge graphs constructed from network logs. To ensure model transparency and interpretability, Local Interpretable Model-Agnostic Explanations (LIME) are incorporated, while structured prompts guide report generation using Generative AI. Experimental results obtained on benchmark datasets demonstrate that the methodology achieves a classification accuracy exceeding 84 %, outperforming existing detection techniques. Additionally, a comprehensive evaluation involving ablation analysis, LLM-based assessments, and expert reviews shows that incorporating structured knowledge and explainability significantly enhances the clarity, correctness, and informativeness of generated reports. These findings confirm the system’s effectiveness both as a detection mechanism and as a practical tool that helps analysts understand threats and craft informed responses.

The proliferation of IoT devices necessitates privacy-preserving on-device training, yet existing Federated Learning systems struggle with privacy leakage and unverifiable aggregation under resource constraints. In this paper, we present DVFL, a lightweight verifiable FL system with dynamic obfuscation tailored for IoT. Distinct from traditional FL, DVFL adopts a decoupled personalization strategy: it trains a globally shared feature extractor while keeping the privacy-sensitive classification head strictly local, replacing it with a normalized dummy value during uplink. This design, supported by a 2-of-2 helper module for zero-sum masking, inherently prevents leakage from sensitive layers while enabling personalized adaptation. Furthermore, we introduce a lightweight dual-verification mechanism that allows clients to independently validate aggregation integrity using low-cost hash operations, avoiding prohibitive digital signatures. Experimental results on MNIST, CIFAR-10, UCI HAR and MIT BIH show that DVFL reduces Membership Inference Attack AUC to ≈ 50% and thwarts gradient leakage, while achieving personalized accuracy comparable to undefended baselines.

As cyber–physical systems continue to increase in complexity, multivariate time series exhibit not only intricate temporal patterns within individual variables but also complex inter-variable dependencies, including both synchronous and asynchronous propagation. Although existing Graph Neural Network (GNN)-based methods perform well in modeling variable dependencies, they often decouple temporal pattern processing from the graph construction. This separation often results in graph structures that lack temporal semantics and struggle to capture delayed responses among variables. To address this, we propose a Position Encoding–Enhanced Adaptive Expansion Graph Attention Network (PEE-AEGAT) for multivariate time series anomaly detection and root cause analysis. First, to embed temporal semantics directly into the graph construction process, we design a Multi-Frequency Positional Encoding Projection (MF-PEP) module that explicitly injects multi-scale periodic patterns and positional information into node embeddings. Second, we introduce a Hybrid Adaptive Graph Structure Learning (HAGSL) strategy that constructs a dual-mode graph by integrating static feature similarity with dynamic delay-aware similarity, together with a density-aware adaptive expansion mechanism for flexibly capturing dependencies. Finally, we develop a delay-aware graph attention network for information aggregation and propose an influence-based scoring mechanism to accurately identify anomaly root causes. Extensive experiments on four public real-world datasets demonstrate that PEE-AEGAT significantly outperforms state-of-the-art baselines in anomaly detection accuracy and root cause localization performance.

Embedding graphs in continuous spaces is a key factor for automatic information extraction in diverse tasks (e.g., learning, inferring, predicting). The reliability of graph embeddings directly depends on how much the geometry of the manifold in continuous space matches the graph structure. State-of-the-art of manifold-based graph embedding algorithms assume that the projection on a tangential space of each point in the manifold (corresponding to a node in the graph) would locally resemble a Euclidean space. Although this condition helps in achieving efficient analytical solutions to the embedding problem, it is not an adequate set-up to work with modern real life graphs, that are characterized by weighted connections across nodes often computed over sparse datasets with missing records. In this work, we introduce a new class of manifold, named soft manifold, that can solve this situation. Soft manifolds are mathematical structures with spherical symmetry where the tangent spaces to each point are hypocycloids whose shape is defined according to the velocity of information propagation across the data points. Experimental results on reconstruction tasks on synthetic and real datasets show how the proposed approach enable more accurate and reliable characterization of graphs in continuous spaces with respect to the state-of-the-art.

This article proposes a new cyberattack on decentralized federated learning (DFL), named user isolation poisoning (UIP). While following the standard DFL protocol of receiving and aggregating benign local models, a malicious user strategically generates and distributes compromised updates to undermine the learning process. The objective of the new UIP attack is to diminish the impact of benign users by isolating their model updates, thereby manipulating the shared model to reduce the learning accuracy. To realize this attack, we design a novel threat model that leverages an adversarial message-passing graph (MPG) neural network. Through iterative message passing, the adversarial MPG progressively refines the representations (also known as embeddings or hidden states) of each benign local model update. By orchestrating feature exchanges among connected nodes in a targeted manner, the malicious users effectively curtail the genuine data features of benign local models, thereby diminishing their overall influence within the DFL process. The MPG-based UIP attack is implemented in PyTorch, demonstrating that it effectively reduces the test accuracy of DFL by 49.5% and successfully evades existing cosine similarity- and Euclidean distance-based defense strategies.

Recent advancements in Graph Neural Networks (GNN) have facilitated their widespread adoption in various applications, including recommendation systems. GNNs have proven to be effective in addressing the challenges posed by recommendation systems by efficiently modeling graphs in which nodes represent users or items and edges denote preference relationships. However, current GNN techniques represent nodes by means of a single static vector, which may inadequately capture the intricate complexities of users and items. To overcome these limitations, we propose a solution integrating a cutting-edge model inspired by category theory: Sheaf4Rec. Unlike single vector representations, Sheaf Neural Networks and their corresponding Laplacians represent each node (and edge) using a vector space. Our approach takes advantage of this theory and results in a more comprehensive representation that can be effectively exploited during inference, providing a versatile method applicable to a wide range of graph-related tasks and demonstrating unparalleled performance. Our proposed model exhibits a noteworthy relative improvement of up to 8.53% on F1-Score@10 and an impressive increase of up to 11.29% on NDCG@10, outperforming existing state-of-the-art models such as Neural Graph Collaborative Filtering (NGCF), KGTORe and other recently developed GNN-based models. In addition to its superior predictive capabilities, Sheaf4Rec shows remarkable improvements in terms of efficiency: we observe substantial runtime improvements ranging from 2.5% up to 37% when compared to other GNN-based competitor models, indicating a more efficient way of handling information while achieving better performance. Code is available at https://github.com/antoniopurificato/Sheaf4Rec .

With the rapid advancement of information technology, traditional centralized cloud computing systems face challenges in meeting the stringent low-latency demands of emerging applications. To tackle this issue, this paper proposes a delay-aware cloud-edge architecture that incorporates the distributed characteristics of edge infrastructure, enabling low latency collaboration among edge nodes within the same geographic region. Furthermore, based on this architecture, a dynamic data replica management scheme is introduced, involving synergistic mechanisms between edge nodes and cloud centers to optimally place data replicas on the most suitable edge nodes. The scheme adopts a hierarchical strategy: edge nodes perform short-term localized management of data replicas, while the cloud executes long-term holistic oversight. Experimental results demonstrate that the dynamic approach effectively reduces user access latency, minimizes replica migration frequency, and decreases network bandwidth consumption.

Neural networks remain black-box systems, unsure about their outputs, and their performance may drop unpredictably in real applications. An open question is how to qualitatively extend neural networks, so that they are sure about their reasoning results, or reasoning-for-sure. Here, we introduce set-theoretic relations explicitly and seamlessly into neural networks by extending vector embedding into sphere embedding, so that part-whole relations can explicitly encode set-theoretic relations through sphere boundaries in the vector space. A reasoning-for-sure neural network successfully constructs, within a constant number M of epochs, a sphere configuration as its semantic model for any consistent set-theoretic relation. We implement Hyperbolic Sphere Neural Network (HSphNN), the first reasoning-for-sure neural network for all types of Aristotelian syllogistic reasoning. Its construction process is realised as a sequence of neighbourhood transitions from the current towards the target configuration. We prove M=1 for HSphNN. In experiments, HSphNN achieves the symbolic level rigour of syllogistic reasoning and successfully checks both decisions and explanations of ChatGPT (gpt-3.5-turbo and gpt-4o) without errors. Through prompts, HSphNN improves the performance of gpt-3.5-turbo from 46.875% to 58.98%, and of gpt-4o from 82.42% to 84.76%. We show ways to extend HSphNN for various kinds of logical and Bayesian reasoning, and to integrate it with traditional neural networks seamlessly.

Hypergraphs provide a flexible framework for modeling high-order (complex) interactions among multiple entities, extending beyond traditional pairwise correlations in graph structures. However, deep hypergraph neural networks (HGNNs) often face the challenge of oversmoothing with increasing depth, similar to issues in graph neural networks (GNNs). While oversmoothing in GNNs has been extensively studied, its implications in relation to hypergraphs are less explored. This paper addresses this gap by first theoretically exploring the reasons behind oversmoothing in deep HGNNs. Our novel insights suggest that a spectral-based hypergraph convolution, equipped with both low-pass and high-pass filters, can potentially mitigate these effects. Motivated by these findings, we introduce FrameHGNN, a framework that utilizes framelet-based hypergraph convolutions integrating tight framelet transforms with both low-pass and high-pass components, as well as the commonly used strategies in designing deep GNN architecture: initial residual and identity mappings. The experiment results on diverse benchmark datasets demonstrate that FrameHGNN outperforms several state-of-the-art models, effectively reducing oversmoothing while improving predictive accuracy. Our contributions not only advance the theoretical understanding of deep hypergraph learning but also provide a practical spectral-based approach for HGNNs, emphasizing the design of multifrequency channels.

Hypergraph neural networks (HNNs) have shown promise in handling tasks characterized by high-order correlations, achieving notable success across various applications. However, there has been limited focus on heterophilic hypergraph learning (HHL), in contrast to the increasing attention given to graph neural networks designed for graphs exhibiting heterophily. This paper aims to pave the way for HHL by addressing key gaps from multiple perspectives: measurement, dataset diversity, and baseline model development. First, we introduce metrics to quantify heterophily in hypergraphs, providing a numerical basis for assessing the homophily/heterophily ratio. Second, we develop diverse benchmark datasets across various real-world scenarios, facilitating comprehensive evaluations of existing HNNs and advancing research in HHL. Additionally, as a novel baseline model, we propose HyperUFG, a framelet-based HNN integrating both low-pass and high-pass filters. Extensive experiments conducted on synthetic and benchmark datasets highlight the challenges current HNNs face with heterophilic hypergraphs, while showcasing that HyperUFG performs competitively and often outperforms many existing models in such scenarios. Overall, our study underscores the urgent need for further exploration and development in this emerging field, with the potential to inspire and guide future research in HHL.

Letting AI agents interact in multi-agent applications adds a layer of complexity to the interpretability and prediction of AI outcomes, with profound implications for their trustworthy adoption in research and society. Game theory offers powerful models to capture and interpret strategic interaction among agents, but requires the support of reproducible, standardized and user-friendly IT frameworks to enable comparison and interpretation of results. To this end, we present FAIRGAME, a Framework for AI Agents Bias Recognition using Game Theory. We describe its implementation and usage, and we employ it to uncover biased outcomes in popular games among AI agents, depending on the employed Large Language Model (LLM) and used language, as well as on the personality trait or strategic knowledge of the agents. Overall, FAIRGAME allows users to reliably and easily simulate their desired games and scenarios and compare the results across simulation campaigns and with game-theoretic predictions, enabling the systematic discovery of biases, the anticipation of emerging behavior out of strategic interplays, and empowering further research into strategic decision-making using LLM agents.

We propose a novel Stochastic Differential Equation (SDE) framework to address the problem of learning uncertainty-aware representations for graph-structured data. While Graph Neural Ordinary Differential Equations (GNODEs) have shown promise in learning node representations, they lack the ability to quantify uncertainty. To address this, we introduce Latent Graph Neural Stochastic Differential Equations (LGNSDE), which enhance GNODE by embedding randomness through a Bayesian prior-posterior mechanism for epistemic uncertainty and Brownian motion for aleatoric uncertainty. By leveraging the existence and uniqueness of solutions to graph-based SDEs, we prove that the variance of the latent space bounds the variance of model outputs, thereby providing theoretically sensible guarantees for the uncertainty estimates. Furthermore, we show mathematically that LGNSDEs are robust to small perturbations in the input, maintaining stability over time. Empirical results across several benchmarks demonstrate that our framework is competitive in out-of-distribution detection, robustness to noise perturbations, and active learning, underscoring the ability of LGNSDEs to quantify uncertainty reliably.

Generative models see increasing use in computer-aided drug design. However, while performing well at capturing distributions of molecular motifs, they often produce synthetically inaccessible molecules. To address this, we introduce SynFlowNet, a GFlowNet model whose action space uses chemical reactions and buyable reactants to sequentially build new molecules. By incorporating forward synthesis as an explicit constraint of the generative mechanism, we aim at bridging the gap between in silico molecular generation and real world synthesis capabilities. We evaluate our approach using synthetic accessibility scores and an independent retrosynthesis tool to assess the synthesizability of our compounds, and motivate the choice of GFlowNets through considerable improvement in sample diversity compared to baselines. Additionally, we identify challenges with reaction encodings that can complicate traversal of the MDP in the backward direction. To address this, we introduce various strategies for learning the GFlowNet backward policy and thus demonstrate how additional constraints can be integrated into the GFlowNet MDP framework. This approach enables our model to successfully identify synthesis pathways for previously unseen molecules.

Computational RNA design tasks are often posed as inverse problems, where sequences are designed based on adopting a single desired secondary structure without considering 3D conformational diversity. We introduce gRNAde, a geometric RNA design pipeline operating on 3D RNA backbones to design sequences that explicitly account for structure and dynamics. gRNAde uses a multi-state Graph Neural Network and autoregressive decoding to generates candidate RNA sequences conditioned on one or more 3D backbone structures where the identities of the bases are unknown. On a single-state fixed backbone re-design benchmark of 14 RNA structures from the PDB identified by Das et al. (2010), gRNAde obtains higher native sequence recovery rates (56% on average) compared to Rosetta (45% on average), taking under a second to produce designs compared to the reported hours for Rosetta. We further demonstrate the utility of gRNAde on a new benchmark of multi-state design for structurally flexible RNAs, as well as zero-shot ranking of mutational fitness landscapes in a retrospective analysis of a recent ribozyme. Experimental wet lab validation on 10 different structured RNA backbones finds that gRNAde has a success rate of 50% at designing pseudoknotted RNA structures, a significant advance over 35% for Rosetta. Open source code and tutorials are available at: github.com/chaitjo/geometric-rna-design

The growing demand for personalized learning underscores the importance of accurately predicting students’ future performance to support tailored education and optimize instructional strategies. Traditional approaches predominantly focus on temporal modeling using historical response records and learning trajectories. While effective, these methods often fall short in capturing the intricate interactions between students and learning content, as well as the subtle semantics of these interactions. To address these gaps, we present EduLLM, the first framework to leverage large language models in combination with hypergraph learning for student performance prediction. The framework incorporates FraS-HNN ($\underline{\mbox{Fra}}$melet-based $\underline{\mbox{S}}$igned $\underline{\mbox{H}}$ypergraph $\underline{\mbox{N}}$eural $\underline{\mbox{N}}$etworks), a novel spectral-based model for signed hypergraph learning, designed to model interactions between students and multiple-choice questions. In this setup, students and questions are represented as nodes, while response records are encoded as positive and negative signed hyperedges, effectively capturing both structural and semantic intricacies of personalized learning behaviors. FraS-HNN employs framelet-based low-pass and high-pass filters to extract multi-frequency features. EduLLM integrates fine-grained semantic features derived from LLMs, synergizing with signed hypergraph representations to enhance prediction accuracy. Extensive experiments conducted on multiple educational datasets demonstrate that EduLLM significantly outperforms state-of-the-art baselines, validating the novel integration of LLMs with FraS-HNN for signed hypergraph learning.

The importance of higher-order relations is widely recognized in numerous real-world systems. However, annotating them is a tedious and sometimes even impossible task. Consequently, current approaches for data modelling either ignore the higher-order interactions altogether or simplify them into pairwise connections. To facilitate higher-order processing, even when a hypergraph structure is not available, we introduce SPHINX, a model that learns to infer a latent hypergraph structure in an unsupervised way, solely from the final task-dependent signal. To ensure broad applicability, we design the model to be end-to-end differentiable, capable of generating a discrete hypergraph structure compatible with any modern hypergraph networks, and easily optimizable without requiring additional regularization losses. Through extensive ablation studies and experiments conducted on four challenging datasets, we demonstrate that our model is capable of inferring suitable latent hypergraphs in both transductive and inductive tasks. Moreover, the inferred latent hypergraphs are interpretable and contribute to enhancing the final performance, outperforming existing methods for hypergraph prediction.

Protein’s backbone flexibility is a crucial property that heavily influences its functionality. Recent work in the field of protein diffusion probabilistic modelling has leveraged Normal Mode Analysis (NMA) and, for the first time, introduced information about large scale protein motion into the generative process. However, obtaining molecules with both the desired dynamics and designable quality has proven challenging. In this work, we present NMA-tune, a new method that introduces the dynamics information to the protein design stage. NMA-tune uses a trainable component to condition the backbone generation on the lowest normal mode of oscillation. We implement NMA-tune as a plug-and-play extension to RFdiffusion, show that the proportion of samples with high quality structure and the desired dynamics is improved as compared to other methods without the trainable component, and we show the presence of the targeted modes in the Molecular Dynamics simulations.

Active Feature Acquisition is an instance-wise, sequential decision making problem. The aim is to dynamically select which feature to measure based on current observations, independently for each test instance. Common approaches either use Reinforcement Learning, which experiences training difficulties, or greedily maximize the conditional mutual information of the label and unobserved features, which makes myopic acquisitions. To address these shortcomings, we introduce a latent variable model, trained in a supervised manner. Acquisitions are made by reasoning about the features across many possible unobserved realizations in a stochastic latent space. Extensive evaluation on a large range of synthetic and real datasets demonstrates that our approach reliably outperforms a diverse set of baselines.

Large Language Models (LLMs) have shown promise as robotic planners but often struggle with long-horizon and complex tasks, especially in specialized environments requiring external knowledge. While hierarchical planning and Retrieval-Augmented Generation (RAG) address some of these challenges, they remain insufficient on their own and a deeper integration is required for achieving more reliable systems. To this end, we propose a neuro-symbolic approach that enhances LLMs-based planners with Knowledge Graph-based RAG for hierarchical plan generation. This method decomposes complex tasks into manageable subtasks, further expanded into executable atomic action sequences. To ensure formal correctness and proper decomposition, we integrate a Symbolic Validator, which also functions as a failure detector by aligning expected and observed world states. Our evaluation against baseline methods demonstrates the consistent significant advantages of integrating hierarchical planning, symbolic verification, and RAG across tasks of varying complexity and different LLMs. Additionally, our experimental setup and novel metrics not only validate our approach for complex planning but also serve as a tool for assessing LLMs’ reasoning and compositional capabilities. Code available at https://github.com/corneliocristina/HVR.

We present a novel, and effective, approach to achieve optimal mesh relocation in finite element methods (FEMs). The cost and accuracy of FEMs is critically dependent on the choice of mesh points. Mesh relocation (r-adaptivity) seeks to optimise the mesh geometry to obtain the best solution accuracy at given computational budget. Classical r-adaptivity relies on the solution of a separate nonlinear “meshing” PDE to determine mesh point locations. This incurs significant cost at remeshing, and relies on estimates that relate interpolation- and FEM-error. Recent machine learning approaches have focused on the construction of fast surrogates for such classical methods. Instead, our new approach trains a graph neural network (GNN) to determine mesh point locations by directly minimising the FE solution error from the PDE system Firedrake to achieve higher solution accuracy. Our GNN architecture closely aligns the mesh solution space to that of classical meshing methodologies, thus replacing classical estimates for optimality with a learnable strategy. This allows for rapid and robust training and results in an extremely efficient and effective GNN approach to online r-adaptivity. Our method outperforms both classical, and prior ML, approaches to r-adaptive meshing. In particular, it achieves lower FE solution error, whilst retaining the significant speed-up over classical methods observed in prior ML work.

Graph neural networks process information on graphs represented at a given resolution scale. We analyze the effect of using various coarse-grained graph resolutions, obtained through the Laplacian renormalization group theory, on node classification tasks. At the core of the theory is the grouping of nodes connected by significant information flow at a given time scale. Representations of the graph at different scales encode interaction information at different ranges. We specifically experiment using representations at the characteristic scale of the graph’s mesoscopic structures. We provide the models with the original graph and the graph represented at the characteristic resolution scale and compare them to models that can only access the original graph. Our results showed that models with access to both the original graph and the characteristic scale graph can achieve statistically significant improvements in test accuracy.

Split learning is a new distributed deep learning method, which facilitates resource-constrained clients to complete training of neural network models. In the model privacy security of split learning, many scholars have proposed many research schemes, but there may be many or direct or indirect privacy problems in the training process, any of which may cause serious privacy disclosure. This paper mainly considers two easily exposed data in the training process of split learning, namely the output data of the cut layer and the label of the data. In these two parts of the data, different algorithms of differential privacy are used to avoid the direct exposure of the data and protect the model privacy.

Hypergraphs effectively model higher-order relationships in natural phenomena, capturing complex interactions beyond pairwise connections. We introduce a novel hypergraph message passing framework inspired by interacting particle systems, where hyperedges act as fields inducing shared node dynamics. By incorporating attraction, repulsion, and Allen-Cahn forcing terms, particles of varying classes and features achieve class-dependent equilibrium, enabling separability through the particle-driven message passing. We investigate both first-order and second-order particle system equations for modeling these dynamics, which mitigate over-smoothing and heterophily thus can capture complete interactions. The more stable second-order system permits deeper message passing. Furthermore, we enhance deterministic message passing with stochastic element to account for interaction uncertainties. We prove theoretically that our approach mitigates over-smoothing by maintaining a positive lower bound on the hypergraph Dirichlet energy during propagation and thus to enable hypergraph message passing to go deep. Empirically, our models demonstrate competitive performance on diverse real-world hypergraph node classification tasks, excelling on both homophilic and heterophilic datasets. Source code is available at \href{https://github.com/Xuan-Elfin/HAMP}{the link}.

Electrocardiogram signals play a pivotal role in cardiovascular diagnostics, providing essential information on electrical hearth activity. However, inherent noise and limited resolution can hinder an accurate interpretation of the recordings. In this paper an advanced Denoising Convolutional Autoencoder designed to process electrocardiogram signals, generating super-resolution reconstructions is proposed; this is followed by in-depth analysis of the enhanced signals. The autoencoder receives a signal window (of 5 s) sampled at 50 Hz (low resolution) as input and reconstructs a denoised super-resolution signal at 500 Hz. The proposed autoencoder is applied to publicly available datasets, demonstrating optimal performance in reconstructing high-resolution signals from very low-resolution inputs sampled at 50 Hz. The results were then compared with current state-of-the-art for electrocardiogram super-resolution, demonstrating the effectiveness of the proposed method. The method achieves a signal-to-noise ratio of 12.20 dB, a mean squared error of 0.0044, and a root mean squared error of 4.86%, which significantly outperforms current state-of-the-art alternatives. This framework can effectively enhance hidden information within signals, aiding in the detection of heart-related diseases. • We defined a novel architecture based on autocencoders which is able to denoise and reconstruct high resolution copies of input low resolution ECG signals. • This unique approach that has not been previously applied to ECG signals. • We also present a deep validation of our approach against traditional and contemporary methods in terms of signal-to-noise ratio, mean squared error, and root mean squared error with those of other widely used ECG signal processing techniques. • The results consistently showed superior performance, further validating the effectiveness of our approach. • Given the increasing reliance on effective and efficient diagnostic techniques in medical practice, especially in cardiology, the findings of our study have significant practical implications.

MOTIVATION: Computational models are crucial for addressing critical questions about systems evolution and deciphering system connections. The pivotal feature of making this concept recognizable from the biological and clinical community is the possibility of quickly inspecting the whole system, bearing in mind the different granularity levels of its components. This holistic view of system behaviour expands the evolution study by identifying the heterogeneous behaviours applicable, e.g. to the cancer evolution study. RESULTS: To address this aspect, we propose a new modelling paradigm, UnifiedGreatMod, which allows modellers to integrate fine-grained and coarse-grained biological information into a unique model. It enables functional studies by combining the analysis of the system's multi-level stable states with its fluctuating conditions. This approach helps to investigate the functional relationships and dependencies among biological entities. This is achieved, thanks to the hybridization of two analysis approaches that capture a system's different granularity levels. The proposed paradigm was then implemented into the open-source, general modelling framework GreatMod, in which a graphical meta-formalism is exploited to simplify the model creation phase and R languages to define user-defined analysis workflows. The proposal's effectiveness was demonstrated by mechanistically simulating the metabolic output of Escherichia coli under environmental nutrient perturbations and integrating a gene expression dataset. Additionally, the UnifiedGreatMod was used to examine the responses of luminal epithelial cells to Clostridium difficile infection. AVAILABILITY AND IMPLEMENTATION: GreatMod https://qbioturin.github.io/epimod/, epimod_FBAfunctions https://github.com/qBioTurin/epimod_FBAfunctions, first case study E. coli https://github.com/qBioTurin/Ec_coli_modelling, second case study C. difficile https://github.com/qBioTurin/EpiCell_CDifficile.

Secure multiparty computing (MPC) encrypts local models containing sensitive information before aggregation and transmits them to centralized servers, effectively preventing direct information leakage and providing an efficient encryption scheme for distributed systems. However, the server may still indirectly obtain local data through methods such as deep leakage from gradients (DLG). In addition, during the process of data transmission between the server and each client, it is easy to be attacked and tampered with by malicious third parties, resulting in biased data obtained by the client and causing training to not converge. To further enhance security and ensure the correct transmission of data, our research group proposes a "2+2" scheme that uses double decomposition and double aggregation to keep the global model hidden from the server and enable the client to self-verify the transmission of data. Specifically, based on the traditional MPC decomposition, each decomposition share is divided into public and private parts. This dual decomposition preserves the lossless encryption and decryption advantages of standard MPC and ensures that the central server obtains a nonreal global model. Moreover, in terms of aggregation, in addition to aggregating the global model, further aggregation of validation parameters is added to enable clients to locally validate the correctness of the aggregated parameters. Theoretical analysis and experimental results show that this improvement scheme significantly enhances the security and privacy protection of training data with minimal communication, computing, and storage overhead.

A heterogeneous information network (heterogeneous graph) federated learning plays a crucial role in enabling multiparty collaboration in the Internet of Things system. However, due to differences in business and data, the local models of each participant are heterogeneous and unable to achieve federated aggregation. Furthermore, the nonindependent and identically distributed (non-IID) coupling topology structure among participants severely impacts the performance of federated learning. Given the lack of appropriate solutions to these issues, this study proposes a novel heterogeneous graph federated learning framework (HGFL+) based on self-simulation and meta-model aggregation, which includes the following two innovative techniques: 1) the missing coupling supplement module simulates new neighbor nodes on its original heterogeneous graph, and constructs associated edges using multiple encoder-decoder structures, thereby achieving the supplement of missing neighbors with better results than external generative methods and 2) the heterogeneous model aggregation algorithm realizes the fusion of multiparty heterogeneous graph information through mapping, splitting, aggregating, and recombining multiple stages based on the meta-model (the largest basic model unit among participants). We theoretically analyzed the applicability and effectiveness of HGFL+, demonstrating the generalization boundary of HGFL+. Meanwhile, multidimensional empirical verification of classification performance, convergence effect, time overhead, model size, and application extension (model, task, domain) validates the effectiveness of the proposed method.

This study addresses the increasing threat to Psychological Well-Being (PWB) posed by Depression, Anxiety, and Stress conditions. Machine learning methods have shown promising results for several psychological conditions. However, the lack of transparency in existing models impedes practical application. The study aims to develop explainable machine learning models for depression, anxiety and stress prediction, focusing on features extracted from tasks involving handwriting and drawing. Two hundred patients completed the Depression, Anxiety, and Stress Scale (DASS-21) and performed seven tasks related to handwriting and drawing. Extracted features, encompassing pressure, stroke pattern, time, space, and pen inclination, were used to train the explainable-by-design Entropy-based Logic Explained Network (e-LEN) model, employing first-order logic rules for explanation. Performance comparison was performed with XGBoost, enhanced by the SHAP explanation method. The trained models achieved notable accuracy in predicting depression (0.749 ±0.089), anxiety (0.721 ±0.088), and stress (0.761 ±0.086) through 10-fold cross-validation (repeated 20 times). The e-LEN model’s logic rules facilitated clinical validation, uncovering correlations with existing clinical literature. While performance remained consistent for depression and anxiety on an independent test dataset, a slight degradation was observed for stress prediction in the test task.

Antibodies are indispensable components of the immune system, known for their specific binding to antigens. Beyond their natural immunological functions, they are fundamental in developing vaccines and therapeutic interventions for infectious diseases. The complex architecture of antibodies, particularly their variable regions responsible for antigen recognition, presents significant challenges for computational modeling. Recent advancements in deep learning have markedly improved protein structure prediction; however, accurately modeling antibody-antigen (Ab-Ag) interactions remains challenging due to the inherent flexibility of antibodies and the dynamic nature of binding processes. In this study, we examine the use of predicted Local Distance Difference Test (pLDDT) scores as indicators of residue and side-chain flexibility to model Ab-Ag interactions through a fingerprint-based approach. We demonstrate the significance of flexibility in different antibody-specific tasks, enhancing the predictive accuracy of Ab-Ag interaction models by 4%, resulting in an AUC-ROC of 92%. In addition, we showcase state-of-the-art performance in paratope prediction. These results emphasize the importance of accounting for conformational flexibility in modeling antibody-antigen interactions and show that pLDDT can serve as a coarse proxy for these dynamic features. By optimizing antibody flexibility using pLDDT, they can be engineered to improve affinity or breadth for a specific target. This approach is particularly beneficial for addressing highly variable pathogens like HIV and SARS-CoV-2, as greater flexibility enhances tolerance to sequence variations in target antigens.

In federated learning, malicious attackers may control clients and servers to perform gradient poisoning, forge aggregation results, and infer individual gradient privacy, posing serious security threats. However, existing research has not effectively addressed these three security requirements under a strong threat model. To tackle this issue, we propose an Efficient Aggregation for Federated Learning with Robustness, Verifiability, and Privacy (EARVP): (1) The Privacy-Preserving Two-Party Kernel Principal Component Analysis (PPTKPCA) combined with the DP-Tolerant Two-Party Density Clustering (DPTTDC) achieves strong robustness; (2) The Distributed Trust Aggregation Integrity Verification (DTAIV) ensures strong verifiability even in the presence of collusion; (3) The Gradient-Lossless Enhancement of Client-Level Differential Privacy (GLECLDP) ensures that the lossless gradient generation stage satisfies malicious privacy security and that gradient updates meet (ϵ, δ)-DP during the defense stage; (4) The entire process employs lightweight protocols to achieve efficiency. Theoretical analysis proves that EARVP ensures semi-honest privacy security, malicious privacy security, and aggregation verifiability. Experimental results further demonstrate the robustness and efficiency of the system. Compared to state-of-the-art algorithms, EARVP improves test accuracy by 14.51%, detection accuracy by 13.37%, reduces poisoning success rate by 1.89%, lowers defense overhead by 13.78% compared to homomorphic encryption schemes, and reduces verification costs by a large magnitude.

Fairness in Federated Learning (FL) is imperative not only for the ethical utilization of technology but also for ensuring that models provide accurate, equitable, and beneficial outcomes across varied user demographics and equipment. This paper proposes a new adversarial architecture, referred to as Adversarial Graph Attention Network (AGAT), which deliberately instigates fairness attacks with an aim to bias the learning process across the FL. The proposed AGAT is developed to synthesize malicious, biasing model updates, where the minimum of Kullback-Leibler (KL) divergence between the user's model update and the global model is maximized. Due to a limited set of labeled input-output biasing data samples, a surrogate model is created, which presents the behavior of a complex malicious model update. Moreover, a graph autoencoder (GAE) is designed within the AGAT architecture, which is trained together with sub-gradient descent to reconstruct manipulatively the correlations of the model updates, and maximize the reconstruction loss while keeping the malicious, biasing model updates undetectable. The proposed AGAT attack is implemented in PyTorch, showing experimentally that AGAT successfully increases the minimum value of KL divergence of benign model updates by 60.9% and bypasses the detection of existing defense models. The source code of the AGAT attack is released on GitHub.

The structural similarities between protein sequences and natural languages have led to parallel advancements in deep learning across both domains. While large language models (LLMs) have achieved much progress in the domain of natural language processing, their potential in protein engineering remains largely unexplored. Previous approaches have equipped LLMs with protein understanding capabilities by incorporating external protein encoders, but this fails to fully leverage the inherent similarities between protein sequences and natural languages, resulting in sub-optimal performance and increased model complexity. To address this gap, we present TourSynbio-7B, the first multi-modal large model specifically designed for protein engineering tasks without external protein encoders. TourSynbio-7B demonstrates that LLMs can inherently learn to understand proteins as language. The model is post-trained and instruction fine-tuned on InternLM2-7B using ProteinLM-Dataset, a dataset comprising 17.46 billion tokens of text and protein sequence for self-supervised pretraining and 893K instructions for supervised fine-tuning. TourSynbio7B outperforms GPT-4 on the ProteinLMBench, a benchmark of 944 manually verified multiple-choice questions, with 62.18% accuracy. Leveraging TourSynbio-7B’s enhanced protein sequence understanding capability, we introduce TourSynbioAgent, an innovative framework capable of performing various protein engineering tasks, including mutation analysis, inverse folding, protein folding, and visualization. TourSynbio-Agent integrates previously disconnected deep learning models in the protein engineering domain, offering a unified conversational user interface for improved usability. Finally, we demonstrate the efficacy of TourSynbio-7B and TourSynbio-Agent through two wet lab case studies on vanilla key enzyme modification and steroid compound catalysis. Our results show that this combination facilitates protein engineering tasks in wet labs, leading to higher positive rates, improved mutations, shorter delivery times, and increased automation. The model weights are available at https://huggingface.co/tsynbio/Toursynbio and codes at https://github.com/tsynbio/TourSynbio.

Graph augmentation methods play a crucial role in improving the performance and enhancing generalisation capabilities in Graph Neural Networks (GNNs). Existing graph augmentation methods mainly perturb the graph structures, and are usually limited to pairwise node relations. These methods cannot fully address the complexities of real-world large-scale networks, which often involve higher-order node relations beyond only being pairwise. Meanwhile, real-world graph datasets are predominantly modelled as simple graphs, due to the scarcity of data that can be used to form higher-order edges. Therefore, reconfiguring the higher-order edges as an integration into graph augmentation strategies lights up a promising research path to address the aforementioned issues. In this paper, we present Topological Augmentation (TopoAug), a novel graph augmentation method that builds a combinatorial complex from the original graph by constructing virtual hyperedges directly from the raw data. TopoAug then produces auxiliary node features by extracting information from the combinatorial complex, which are used for enhancing GNN performances on downstream tasks. We design three diverse virtual hyperedge construction strategies to accompany the construction of combinatorial complexes: (1) via graph statistics, (2) from multiple data perspectives, and (3) utilising multi-modality. Furthermore, to facilitate TopoAug evaluation, we provide 23 novel real-world graph datasets across various domains including social media, biology, and e-commerce. Our empirical study shows that TopoAug consistently and significantly outperforms GNN baselines and other graph augmentation methods, across a variety of application contexts, which clearly indicates that it can effectively incorporate higher-order node relations into the graph augmentation for real-world complex networks.

Fact verification aims to verify a claim using evidence from a trustworthy knowledge base. To address this challenge, algorithms must produce features for every claim that are both semantically meaningful, and compact enough to find a semantic alignment with the source information. In contrast to previous work, which tackled the alignment problem by learning over annotated corpora of claims and their corresponding labels, we propose SFAVEL ($\underline{S}$elf-supervised $\underline{Fa}$ct $\underline{Ve}$rification via $\underline{L}$anguage Model Distillation), a novel unsupervised pretraining framework that leverages pre-trained language models to distil self-supervised features into high-quality claim-fact alignments without the need for annotations. This is enabled by a novel contrastive loss function that encourages features to attain high-quality claim and evidence alignments whilst preserving the semantic relationships across the corpora. Notably, we present results that achieve a new state-of-the-art on FB15k-237 (+5.3\% Hits@1) and FEVER (+8\% accuracy) with linear evaluation.

We introduce ProteinWorkshop, a comprehensive benchmark suite for representation learning on protein structures with Geometric Graph Neural Networks. We consider large-scale pre-training and downstream tasks on both experimental and predicted structures to enable the systematic evaluation of the quality of the learned structural representation and their usefulness in capturing functional relationships for downstream tasks. We find that: (1) large-scale pretraining on AlphaFold structures and auxiliary tasks consistently improve the performance of both rotation-invariant and equivariant GNNs, and (2) more expressive equivariant GNNs benefit from pretraining to a greater extent compared to invariant models. We aim to establish a common ground for the machine learning and computational biology communities to rigorously compare and advance protein structure representation learning. Our open-source codebase reduces the barrier to entry for working with large protein structure datasets by providing: (1) storage-efficient dataloaders for large-scale structural databases including AlphaFoldDB and ESM Atlas, as well as (2) utilities for constructing new tasks from the entire PDB. ProteinWorkshop is available at: github.com/a-r-j/ProteinWorkshop.

Current protein generative models are able to design novel backbones with desired shapes or functional motifs. However, despite the importance of a protein’s dynamical properties for its function, conditioning on dynamical properties remains elusive. We present a new approach to protein generative modeling by leveraging Normal Mode Analysis that enables us to capture dynamical properties too. We introduce a method for conditioning the diffusion probabilistic models on protein dynamics, specifically on the lowest non-trivial normal mode of oscillation. Our method, similar to the classifier guidance conditioning, formulates the sampling process as being driven by conditional and unconditional terms. However, unlike previous works, we approximate the conditional term with a simple analytical function rather than an external neural network, thus making the eigenvector calculations approachable. We present the corresponding SDE theory as a formal justification of our approach. We extend our framework to conditioning on structure and dynamics at the same time, enabling scaffolding of the dynamical motifs. We demonstrate the empirical effectiveness of our method by turning the open-source unconditional protein diffusion model Genie into the conditional model with no retraining. Generated proteins exhibit the desired dynamical and structural properties while still being biologically plausible. Our work represents a first step towards incorporating dynamical behaviour in protein design and may open the door to designing more flexible and functional proteins in the future.

Spectral Graph Neural Networks (GNNs), alternatively known as graph filters, have gained increasing prevalence for heterophily graphs. Optimal graph filters rely on Laplacian eigendecomposition for Fourier transform. In an attempt to avert prohibitive computations, numerous polynomial filters have been proposed. However, polynomials in the majority of these filters are predefined and remain fixed across different graphs, failing to accommodate the varying degrees of heterophily. Addressing this gap, we demystify the intrinsic correlation between the spectral property of desired polynomial bases and the heterophily degrees via thorough theoretical analyses. Subsequently, we develop a novel adaptive heterophily basis wherein the basis vectors mutually form angles reflecting the heterophily degree of the graph. We integrate this heterophily basis with the homophily basis to construct a universal polynomial basis UniBasis, which devises a polynomial filter based graph neural network – UniFilter. It optimizes the convolution and propagation in GNN, thus effectively limiting over-smoothing and alleviating over-squashing. Our extensive experiments, conducted on datasets with a diverse range of heterophily, support the superiority of UniBasis in the universality but also its proficiency in graph explanation.

Topological deep learning (TDL) is a rapidly evolving field that uses topological features to understand and design deep learning models. This paper posits that TDL is the new frontier for relational learning. TDL may complement graph representation learning and geometric deep learning by incorporating topological concepts, and can thus provide a natural choice for various machine learning settings. To this end, this paper discusses open problems in TDL, ranging from practical benefits to theoretical foundations. For each problem, it outlines potential solutions and future research opportunities. At the same time, this paper serves as an invitation to the scientific community to actively participate in TDL research to unlock the potential of this emerging field.

Proteins perform much of the work in living organisms, and consequently the development of efficient computational methods for protein representation is essential for advancing large-scale biological research. Most current approaches struggle to efficiently integrate the wealth of information contained in the protein sequence and structure. In this paper, we propose a novel framework for embedding protein graphs in geometric vector spaces, by learning an encoder function that preserves the structural distance between protein graphs. Utilizing Graph Neural Networks (GNNs) and Large Language Models (LLMs), the proposed framework generates structure- and sequence- aware protein representations. We demonstrate that our embeddings are successful in the task of comparing protein structures, while providing a significant speed-up compared to traditional approaches based on structural alignment. Our framework achieves remarkable results in the task of protein structure classification; in particular, when compared to other work, the proposed method shows an average F1-Score improvement of 26% on out-of-distribution (OOD) samples and of 32% when tested on samples coming from the same distribution as the training data. Our approach finds applications in areas such as drug prioritization, drug re-purposing, disease sub-type analysis and elsewhere.

Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) involves acquiring a sequence of MRIs during the administration of a contrast agent. Radiologists then aim to discern the contrast uptake differences between malignant and benign lesions for tumor classification. Regrettably, existing literature underutilizes the temporal structure inherent to DCEMRI time series, leading to tumor classifications based on individual instants rather than entire sequences. This research introduces two Graph Neural Network (GNN)-based methods designed to aggregate information from multiple instants within the DCE-MRI sequence. Each lesion undergoes manual segmentation, and radiomic features are individually extracted from each time instant of the DCE-MRI sequence. Two graph construction methodologies are proposed: (i) a fully connected graph topology, aiming to represent each temporal instant as a node in a graph; (ii) a multiplex network, named MUGI-MRI (MUltiplex Graph neural network for Integration of MRI), where each layer identifies an instant of the DCE-MRI sequence. MUGI-MRI achieves an AUROC of 0.8017 ± 0.1146, showcasing promising performance in lesion classification. In addition to improving upon current state-of-the-art, the integration capability of MUGIMRI addresses the problem of imbalance between sensitivity and specificity, which affects numerous studies in the realm of DCE-MRI. Our findings strongly indicate that the aggregation of information across all time instants is pivotal for enhancing the diagnostic process, and vastly superior to a simplistic instant-wise analysis. While applied to MRI sequences, our approach can be extended to general problems of multimodal data integration.

Graph Neural Networks (GNNs) have demonstrated remarkable success in learning from graph-structured data. However, they face significant limitations in expressive power, struggling with long-range interactions and lacking a principled approach to modeling higher-order structures and group interactions. Cellular Isomorphism Networks (CINs) recently addressed most of these challenges with a message passing scheme based on cell complexes. Despite their advantages, CINs make use only of boundary and upper messages which do not consider a direct interaction between the rings present in the underlying complex. Accounting for these interactions might be crucial for learning representations of many real-world complex phenomena such as the dynamics of supramolecular assemblies, neural activity within the brain, and gene regulation processes. In this work, we propose CIN++, an enhancement of the topological message passing scheme introduced in CINs. Our message passing scheme accounts for the aforementioned limitations by letting the cells receive also lower messages within each layer. By providing a more comprehensive representation of higher-order and long-range interactions, our enhanced topological message passing scheme achieves state-of-the-art results on large-scale and long-range chemistry benchmarks.

ACM Multimedia 2024, Melbourne, Australia, Oct 28 - Nov 1, 2024

Generative modelling paradigms based on denoising diffusion processes have emerged as a leading candidate for conditional sampling in inverse problems. In many real-world applications, we often have access to large, expensively trained unconditional diffusion models, which we aim to exploit for improving conditional sampling. Most recent approaches are motivated heuristically and lack a unifying framework, obscuring connections between them. Further, they often suffer from issues such as being very sensitive to hyperparameters, being expensive to train or needing access to weights hidden behind a closed API. In this work, we unify conditional training and sampling using the mathematically well-understood Doob's h-transform. This new perspective allows us to unify many existing methods under a common umbrella. Under this framework, we propose DEFT (Doob's h-transform Efficient FineTuning), a new approach for conditional generation that simply fine-tunes a very small network to quickly learn the conditional $h$-transform, while keeping the larger unconditional network unchanged. DEFT is much faster than existing baselines while achieving state-of-the-art performance across a variety of linear and non-linear benchmarks. On image reconstruction tasks, we achieve speedups of up to 1.6$\times$, while having the best perceptual quality on natural images and reconstruction performance on medical images. Further, we also provide initial experiments on protein motif scaffolding and outperform reconstruction guidance methods.

Neural Algorithmic Reasoning (NAR) research has demonstrated that graph neural networks (GNNs) could learn to execute classical algorithms. However, most previous approaches have always used a recurrent architecture, where each iteration of the GNN matches an iteration of the algorithm. In this paper we study neurally solving algorithms from a different perspective: since the algorithm’s solution is often an equilibrium, it is possible to find the solution directly by solving an equilibrium equation. Our approach requires no information on the ground-truth number of steps of the algorithm, both during train and test time. Furthermore, the proposed method improves the performance of GNNs on executing algorithms and is a step towards speeding up existing NAR models. Our empirical evidence, leveraging algorithms from the CLRS-30 benchmark, validates that one can train a network to solve algorithmic problems by directly finding the equilibrium. We discuss the practical implementation of such models and propose regularisations to improve the performance of these equilibrium reasoners.

Graph Neural Networks (GNNs), known as spectral graph filters, find a wide range of applications in web networks. To bypass eigendecomposition, polynomial graph filters are proposed to approximate graph filters by leveraging various polynomial bases for filter training. However, no existing studies have explored the diverse polynomial graph filters from a unified perspective for optimization.

Abstract Autism Spectrum Disorder (ASD) is a chronic condition characterised by impairments in social interaction and communication. Early detection of ASD is desired, and there exists a demand for the development of diagnostic aids to facilitate this. A lightweight Involutional Neural Network (INN) architecture has been developed to diagnose ASD. The model follows a simpler architectural design and has less number of parameters than the state‐of‐the‐art (SOTA) image classification models, requiring lower computational resources. The proposed model is trained to detect ASD from eye‐tracking scanpath (SP), heatmap (HM), and fixation map (FM) images. Monte Carlo Dropout has been applied to the model to perform an uncertainty analysis and ensure the effectiveness of the output provided by the proposed INN model. The model has been trained and evaluated using two publicly accessible datasets. From the experiment, it is seen that the model has achieved 98.12% accuracy, 96.83% accuracy, and 97.61% accuracy on SP, FM, and HM, respectively, which outperforms the current SOTA image classification models and other existing works conducted on this topic.

This paper introduces Elegans-AI models, a class of neural networks that leverage the connectome topology of the Caenorhabditis elegans to design deep and reservoir architectures. Utilizing deep learning models inspired by the connectome, this paper leverages the evolutionary selection process to consolidate the functional arrangement of biological neurons within their networks. The initial goal involves the conversion of natural connectomes into artificial representations. The second objective centers on embedding the complex circuitry topology of artificial connectomes into both deep learning and deep reservoir networks, highlighting their neural-dynamic short-term and long-term memory and learning capabilities. Lastly, our third objective aims to establish structural explainability by examining the heterophilic/homophilic properties within the connectome and their impact on learning capabilities. In our study, the Elegans-AI models demonstrate superior performance compared to similar models that utilize either randomly rewired artificial connectomes or simulated bio-plausible ones. Notably, these Elegans-AI models achieve a top-1 accuracy of 99.99% on both Cifar10 and Cifar100, and 99.84% on MNIST Unsup. They do this with significantly fewer learning parameters, particularly when reservoir configurations of the connectome are used. Our findings indicate a clear connection between bio-plausible network patterns, the small-world characteristic, and learning outcomes, emphasizing the significant role of evolutionary optimization in shaping the topology of artificial neural networks for improved learning performance.

Histopathology image-based survival prediction aims to provide a precise assessment of cancer prognosis and can inform personalized treatment decision-making in order to improve patient outcomes. However, existing methods cannot automatically model the complex correlations between numerous morphologically diverse patches in each whole slide image (WSI), thereby preventing them from achieving a more profound understanding and inference of the patient status. To address this, here we propose a novel deep learning framework, termed dual-stream multi-dependency graph neural network (DM-GNN), to enable precise cancer patient survival analysis. Specifically, DM-GNN is structured with the feature updating and global analysis branches to better model each WSI as two graphs based on morphological affinity and global co-activating dependencies. As these two dependencies depict each WSI from distinct but complementary perspectives, the two designed branches of DM-GNN can jointly achieve the multi-view modeling of complex correlations between the patches. Moreover, DM-GNN is also capable of boosting the utilization of dependency information during graph construction by introducing the affinity-guided attention recalibration module as the readout function. This novel module offers increased robustness against feature perturbation, thereby ensuring more reliable and stable predictions. Extensive benchmarking experiments on five TCGA datasets demonstrate that DM-GNN outperforms other state-of-the-art methods and offers interpretable prediction insights based on the morphological depiction of high-attention patches. Overall, DM-GNN represents a powerful and auxiliary tool for personalized cancer prognosis from histopathology images and has great potential to assist clinicians in making personalized treatment decisions and improving patient outcomes.

Graph neural networks (GNNs) are powerful models for processing graph data and have demonstrated state-of-the-art performance on many downstream tasks. However, existing GNNs can generally suffer from two limitations: over-smoothing and over-squashing, which can significantly undermine their learning ability for large graphs. To overcome these issues simultaneously, by utilizing the concept of effective resistances, we focus on minimizing total constrained resistance while identifying problematic edges using topological redundancy and bottleneck sparsity coefficients. We introduce a novel graph rewiring and preprocessing method guided by effective resistance (GPER), capable of edge addition or removal. Theoretical analysis validates our method's efficacy in mitigating over-smoothing and over-squashing. In the experiments, we conduct node and graph classifications on the benchmark datasets and can achieve an average improvement of 7.8% and 2.0%, respectively. We also conduct scalability analysis on large graphs with GCN and demonstrate that the proposed preprocess approach can reduce graph size by over 50% while improve the performance.

Deep neural networks for graphs (DNNGs) represent an emerging field that studies how the deep learning method can be generalized to graph-structured data. Since graphs are a powerful and flexible tool to represent complex information in the form of patterns and their relationships, ranging from molecules to protein-to-protein interaction networks, to social or transportation networks, or up to knowledge graphs, potentially modeling systems at very different scales, these methods have been exploited for many application domains.

Tabular biomedical data is often high-dimensional but with a very small number of samples. Although recent work showed that well-regularised simple neural networks could outperform more sophisticated architectures on tabular data, they are still prone to overfitting on tiny datasets with many potentially irrelevant features. To combat these issues, we propose Weight Predictor Network with Feature Selection (WPFS) for learning neural networks from high-dimensional and small sample data by reducing the number of learnable parameters and simultaneously performing feature selection. In addition to the classification network, WPFS uses two small auxiliary networks that together output the weights of the first layer of the classification model. We evaluate on nine real-world biomedical datasets and demonstrate that WPFS outperforms other standard as well as more recent methods typically applied to tabular data. Furthermore, we investigate the proposed feature selection mechanism and show that it improves performance while providing useful insights into the learning task.

Graph neural networks (GNNs) are highly effective on a variety of graph-related tasks; however, they lack interpretability and transparency. Current explainability approaches are typically local and treat GNNs as black-boxes. They do not look inside the model, inhibiting human trust in the model and explanations. Motivated by the ability of neurons to detect high-level semantic concepts in vision models, we perform a novel analysis on the behaviour of individual GNN neurons to answer questions about GNN interpretability. We propose a novel approach for producing global explanations for GNNs using neuron-level concepts to enable practitioners to have a high-level view of the model. Specifically, (i) to the best of our knowledge, this is the first work which shows that GNN neurons act as concept detectors and have strong alignment with concepts formulated as logical compositions of node degree and neighbourhood properties; (ii) we quantitatively assess the importance of detected concepts, and identify a trade-off between training duration and neuron-level interpretability; (iii) we demonstrate that our global explainability approach has advantages over the current state-of-the-art -- we can disentangle the explanation into individual interpretable concepts backed by logical descriptions, which reduces potential for bias and improves user-friendliness.

Synthetic data is becoming an increasingly promising technology, and successful applications can improve privacy, fairness, and data democratization. While there are many methods for generating synthetic tabular data, the task remains non-trivial and unexplored for specific scenarios. One such scenario is survival data. Here, the key difficulty is censoring: for some instances, we are not aware of the time of event, or if one even occurred. Imbalances in censoring and time horizons cause generative models to experience three new failure modes specific to survival analysis: (1) generating too few at-risk members; (2) generating too many at-risk members; and (3) censoring too early. We formalize these failure modes and provide three new generative metrics to quantify them. Following this, we propose SurvivalGAN, a generative model that handles survival data firstly by addressing the imbalance in the censoring and event horizons, and secondly by using a dedicated mechanism for approximating time-to-event/censoring. We evaluate this method via extensive experiments on medical datasets. SurvivalGAN outperforms multiple baselines at generating survival data, and in particular addresses the failure modes as measured by the new metrics, in addition to improving downstream performance of survival models trained on the synthetic data.

Integrating transformers with graph representation learning has emerged as a research focal point. However, recent studies showed that positional encoding in Transformers does not capture enough structural information between nodes. Additionally, existing graph neural network (GNN) models face the oversquashing issue, impeding information retention from distant nodes. To address, we transform graphs into regular structures, such as tokens, to enhance positional understanding and leverage transformer strengths. Inspired by the visual transformer (ViT) model, we propose partitioning graphs into patches and apply GNN models obtain fixed size vectors. Notably, our approach adopts contrastive learning for in-depth graph structure and incorporate more topological information via Ricci curvature to alleviate over-squashing problem by attenuating the effects of negatively curved edges while preserving the original graph structure. Unlike existing graph rewiring methods that directly modify graph structure by adding or removing edges, this approach is potentially more suitable for applications such as molecular learning where structural preservation is important. Our innovative pipeline subsequently introduces the PerformerMixer, a transformer variant with linear complexity, ensuring efficient computation. Evaluations on real-world benchmarks demonstrate our framework’s superior performance, like Peptides-func and achieve 3-WL expressiveness.

Shadows in videos are difficult to detect because of the large shadow deformation between frames. In this work, we argue that accounting for shadow deformation is essential when designing a video shadow detection method. To this end, we introduce the shadow deformation attention trajectory (SODA), a new type of video self-attention module, specially designed to handle the large shadow deformations in videos. Moreover, we present a new shadow contrastive learning mechanism (SCOTCH) which aims at guiding the network to learn a unified shadow representation from massive positive shadow pairs across different videos. We demonstrate empirically the effectiveness of our two contributions in an ablation study. Furthermore, we show that SCOTCH and SODA significantly outperforms existing techniques for video shadow detection. Code is available at the project page: https://lihaoliu-cambridge.github.io/scotch_and_soda/

Deep learning methods are highly accurate, yet their opaque decision process prevents them from earning full human trust. Concept-based models aim to address this issue by learning tasks based on a set of human-understandable concepts. However, state-of-the-art concept-based models rely on high-dimensional concept embedding representations which lack a clear semantic meaning, thus questioning the interpretability of their decision process. To overcome this limitation, we propose the Deep Concept Reasoner (DCR), the first interpretable concept-based model that builds upon concept embeddings. In DCR, neural networks do not make task predictions directly, but they build syntactic rule structures using concept embeddings. DCR then executes these rules on meaningful concept truth degrees to provide a final interpretable and semantically-consistent prediction in a differentiable manner. Our experiments show that DCR: (i) improves up to +25% w.r.t. state-of-the-art interpretable concept-based models on challenging benchmarks (ii) discovers meaningful logic rules matching known ground truths even in the absence of concept supervision during training, and (iii), facilitates the generation of counterfactual examples providing the learnt rules as guidance.

Message Passing Neural Networks (MPNNs) are instances of Graph Neural Networks that leverage the graph to send messages over the edges. This inductive bias leads to a phenomenon known as over-squashing, where a node feature is insensitive to information contained at distant nodes. Despite recent methods introduced to mitigate this issue, an understanding of the causes for over-squashing and of possible solutions are lacking. In this theoretical work, we prove that: (i) Neural network width can mitigate over-squashing, but at the cost of making the whole network more sensitive; (ii) Conversely, depth cannot help mitigate over-squashing: increasing the number of layers leads to over-squashing being dominated by vanishing gradients; (iii) The graph topology plays the greatest role, since over-squashing occurs between nodes at high commute time. Our analysis provides a unified framework to study different recent methods introduced to cope with over-squashing and serves as a justification for a class of methods that fall under graph rewiring.

The expressive power of Graph Neural Networks (GNNs) has been studied extensively through the Weisfeiler-Leman (WL) graph isomorphism test. However, standard GNNs and the WL framework are inapplicable for geometric graphs embedded in Euclidean space, such as biomolecules, materials, and other physical systems. In this work, we propose a geometric version of the WL test (GWL) for discriminating geometric graphs while respecting the underlying physical symmetries: permutations, rotation, reflection, and translation. We use GWL to characterise the expressive power of geometric GNNs that are invariant or equivariant to physical symmetries in terms of distinguishing geometric graphs. GWL unpacks how key design choices influence geometric GNN expressivity: (1) Invariant layers have limited expressivity as they cannot distinguish one-hop identical geometric graphs; (2) Equivariant layers distinguish a larger class of graphs by propagating geometric information beyond local neighbourhoods; (3) Higher order tensors and scalarisation enable maximally powerful geometric GNNs; and (4) GWL's discrimination-based perspective is equivalent to universal approximation. Synthetic experiments supplementing our results are available at https://github.com/chaitjo/geometric-gnn-dojo

Higher-order relations are widespread in nature, with numerous phenomena involving complex interactions that extend beyond simple pairwise connections. As a result, advancements in higher-order processing can accelerate the growth of various fields requiring structured data. Current approaches typically represent these interactions using hypergraphs. We enhance this representation by introducing cellular sheaves for hypergraphs, a mathematical construction that adds extra structure to the conventional hypergraph while maintaining their local, higher-order connectivity. Drawing inspiration from existing Laplacians in the literature, we develop two unique formulations of sheaf hypergraph Laplacians: linear and non-linear. Our theoretical analysis demonstrates that incorporating sheaves into the hypergraph Laplacian provides a more expressive inductive bias than standard hypergraph diffusion, creating a powerful instrument for effectively modelling complex data structures. We employ these sheaf hypergraph Laplacians to design two categories of models: Sheaf Hypergraph Neural Networks and Sheaf Hypergraph Convolutional Networks. These models generalize classical Hypergraph Networks often found in the literature. Through extensive experimentation, we show that this generalization significantly improves performance, achieving top results on multiple benchmark datasets for hypergraph node classification.

The rise of Artificial Intelligence (AI) recently empowered researchers to investigate hard mathematical problems which eluded traditional approaches for decades. Yet, the use of AI in Universal Algebra (UA)---one of the fields laying the foundations of modern mathematics---is still completely unexplored. This work proposes the first use of AI to investigate UA's conjectures with an equivalent equational and topological characterization. While topological representations would enable the analysis of such properties using graph neural networks, the limited transparency and brittle explainability of these models hinder their straightforward use to empirically validate existing conjectures or to formulate new ones. To bridge these gaps, we propose a general algorithm generating AI-ready datasets based on UA's conjectures, and introduce a novel neural layer to build fully interpretable graph networks. The results of our experiments demonstrate that interpretable graph networks: (i) enhance interpretability without sacrificing task accuracy, (ii) strongly generalize when predicting universal algebra's properties, (iii) generate simple explanations that empirically validate existing conjectures, and (iv) identify subgraphs suggesting the formulation of novel conjectures.

Inverse protein folding is challenging due to its inherent one-to-many mapping characteristic, where numerous possible amino acid sequences can fold into a single, identical protein backbone. This task involves not only identifying viable sequences but also representing the sheer diversity of potential solutions. However, existing discriminative models, such as transformer-based auto-regressive models, struggle to encapsulate the diverse range of plausible solutions. In contrast, diffusion probabilistic models, as an emerging genre of generative approaches, offer the potential to generate a diverse set of sequence candidates for determined protein backbones. We propose a novel graph denoising diffusion model for inverse protein folding, where a given protein backbone guides the diffusion process on the corresponding amino acid residue types. The model infers the joint distribution of amino acids conditioned on the nodes' physiochemical properties and local environment. Moreover, we utilize amino acid replacement matrices for the diffusion forward process, encoding the biologically-meaningful prior knowledge of amino acids from their spatial and sequential neighbors as well as themselves, which reduces the sampling space of the generative process. Our model achieves state-of-the-art performance over a set of popular baseline methods in sequence recovery and exhibits great potential in generating diverse protein sequences for a determined protein backbone structure.

Graph classification aims to categorise graphs based on their structure and node attributes. In this work, we propose to tackle this task using tools from graph signal processing by deriving spectral features, which we then use to design two variants of Gaussian process models for graph classification. The first variant uses spectral features based on the distribution of energy of a node feature signal over the spectrum of the graph. We show that even such a simple approach, having no learned parameters, can yield competitive performance compared to strong neural network and graph kernel baselines. A second, more sophisticated variant is designed to capture multi-scale and localised patterns in the graph by learning spectral graph wavelet filters, obtaining improved performance on synthetic and real-world data sets. Finally, we show that both models produce well calibrated uncertainty estimates, enabling reliable decision making based on the model predictions.

Drug Side-Effects (DSEs) have a high impact on public health, care system costs, and drug discovery processes. Predicting the probability of side-effects, before their occurrence, is fundamental to reduce this impact, in particular on drug discovery. Candidate molecules could be screened before undergoing clinical trials, reducing the costs in time, money, and health of the participants. Drug side-effects are triggered by complex biological processes involving many different entities, from drug structures to protein-protein interactions. To predict their occurrence, it is necessary to integrate data from heterogeneous sources. In this work, such heterogeneous data is integrated into a graph dataset, expressively representing the relational information between different entities, such as drug molecules and genes. The relational nature of the dataset represents an important novelty for drug side-effect predictors. Graph Neural Networks (GNNs) are exploited to predict DSEs on our dataset with very promising results. GNNs are deep learning models that can process graph-structured data, with minimal information loss, and have been applied on a wide variety of biological tasks. Our experimental results confirm the advantage of using relationships between data entities, suggesting interesting future developments in this scope. The experimentation also shows the importance of specific subsets of data in determining associations between drugs and side-effects.

Recent studies on multi-contrast MRI reconstruction have demonstrated the potential of further accelerating MRI acquisition by exploiting correlation between contrasts. Most of the state-of-the-art approaches have achieved improvement through the development of network architectures for fixed under-sampling patterns, without considering inter-contrast correlation in the under-sampling pattern design. On the other hand, sampling pattern learning methods have shown better reconstruction performance than those with fixed under-sampling patterns. However, most under-sampling pattern learning algorithms are designed for single contrast MRI without exploiting complementary information between contrasts. To this end, we propose a framework to optimize the under-sampling pattern of a target MRI contrast which complements the acquired fully-sampled reference contrast. Specifically, a novel image synthesis network is introduced to extract the redundant information contained in the reference contrast, which is exploited in the subsequent joint pattern optimization and reconstruction network. We have demonstrated superior performance of our learned under-sampling patterns on both public and in-house datasets, compared to the commonly used under-sampling patterns and state-of-the-art methods that jointly optimize the reconstruction network and the under-sampling patterns, up to 8-fold under-sampling factor.

After analyzing the long-term evolution (LTE) authentication and key agreement process (EPS-AKA), its existing security vulnerabilities are pointed out. Based on elliptic curve cryptography (ECC) self-certified public keys, this paper proposes an ECC self-certified authentication key agreement scheme (ESC-AKA). This scheme includes the addition of a trusted center (TC), which generates the public keys for the home subscriber server (HSS), the mobility management entity (MME), and the user equipment (UE). Three communication protocols are designed, including MME/HSS registration, UE registration, and UE access. A strand space model is used to carry out the formal analysis, and performance and security analyses are carried out. The results show that this scheme can compensate for the security vulnerabilities of the original EPS-AKA scheme. It implements the encrypted transmission of the international mobile subscriber identity (IMSI), and realizes the mutual authentication between the HSS and MME, the MME and UE, and the HSS and UE. Because the self-certified public key cryptosystem is adopted in this scheme, communication encryption is ensured, and the risk of the TC simultaneously mastering the public and private keys is avoided. This scheme is proven to be effective in protecting the communication security of the LTE network.

With the wide application of deep learning, the amount of data required to train deep learning models is becoming increasingly larger, resulting in an increased training time and higher requirements for computing resources. To improve the throughput of a distributed learning system, task scheduling and resource scheduling are required. This paper proposes to combine ARIMA and GRU models to predict the future task volume. In terms of task scheduling, multi-priority task queues are used to divide tasks into different queues according to their priorities to ensure that high-priority tasks can be completed in advance. In terms of resource scheduling, the reinforcement learning method is adopted to manage limited computing resources. The reward function of reinforcement learning is constructed based on the resources occupied by the task, the training time, the accuracy of the model. When a distributed learning model tends to converge, the computing resources of the task are gradually reduced so that they can be allocated to other learning tasks. The results of experiments demonstrate that RLPTO tends to use more compu-ting nodes when facing tasks with large data scale and has good scalability. The distributed learning system reward experiment shows that RLPTO can make the computing cluster get the largest reward.

Explainable artificial intelligence has rapidly emerged since lawmakers have started requiring interpretable models for safety-critical domains. Concept-based neural networks have arisen as explainable-by-design methods as they leverage human-understandable symbols (i.e. concepts) to predict class memberships. However, most of these approaches focus on the identification of the most relevant concepts but do not provide concise, formal explanations of how such concepts are leveraged by the classifier to make predictions. In this paper, we propose a novel end-to-end differentiable approach enabling the extraction of logic explanations from neural networks using the formalism of First-Order Logic. The method relies on an entropy-based criterion which automatically identifies the most relevant concepts. We consider four different case studies to demonstrate that: (i) this entropy-based criterion enables the distillation of concise logic explanations in safety-critical domains from clinical data to computer vision; (ii) the proposed approach outperforms state-of-the-art white-box models in terms of classification accuracy.

Recent research on graph neural network (GNN) models successfully applied GNNs to classical graph algorithms and combinatorial optimisation problems. This has numerous benefits, such as allowing applications of algorithms when preconditions are not satisfied, or reusing learned models when sufficient training data is not available or can't be generated. Unfortunately, a key hindrance of these approaches is their lack of explainability, since GNNs are black-box models that cannot be interpreted directly. In this work, we address this limitation by applying existing work on concept-based explanations to GNN models. We introduce concept-bottleneck GNNs, which rely on a modification to the GNN readout mechanism. Using three case studies we demonstrate that: (i) our proposed model is capable of accurately learning concepts and extracting propositional formulas based on the learned concepts for each target class; (ii) our concept-based GNN models achieve comparative performance with state-of-the-art models; (iii) we can derive global graph concepts, without explicitly providing any supervision on graph-level concepts.

Link prediction aims to reveal missing edges in a graph. We introduce a deep graph convolutional Gaussian process model for this task, which addresses recent challenges in graph machine learning with oversmoothing and overfitting. Using simplified graph convolutions, we transform a Gaussian process to leverage the topological information of the graph domain. To scale the Gaussian process model to larger graphs, we introduce a variational inducing point method that places pseudo-inputs on a graph-structured domain. Multiple Gaussian processes are assembled into a hierarchy whose structure allows skipping convolutions and thus counteracting oversmoothing. The proposed model represents the first Gaussian process for link prediction that makes use of both node features and topological information. We evaluate our model on multiple graph data sets with up to thousands of nodes and report consistent improvements over competitive link prediction approaches.

Graph-based models require aggregating information in the graph from neighbourhoods of different sizes. In particular, when the data exhibit varying levels of smoothness on the graph, a multi-scale approach is required to capture the relevant information. In this work, we propose a Gaussian process model using spectral graph wavelets, which can naturally aggregate neighbourhood information at different scales. Through maximum likelihood optimisation of the model hyperparameters, the wavelets automatically adapt to the different frequencies in the data, and as a result our model goes beyond capturing low frequency information. We achieve scalability to larger graphs by using a spectrum-adaptive polynomial approximation of the filter function, which is designed to yield a low approximation error in dense areas of the graph spectrum. Synthetic and real-world experiments demonstrate the ability of our model to infer scales accurately and produce competitive performances against state-of-the-art models in graph-based learning tasks.

The internal clock of the CPU uses oscillators made from quartz crystals. Small changes in these crystals can cause small but measurable differences in the clock frequency. Under a low CPU load, the function execution time distribution follows the Pareto distribution. However, the function execution time distribution no longer follows the Pareto distribution when the CPU load is high, and the CPU clock fingerprint becomes invalid. In view of this problem, this paper proposes an adaptive Pareto principle that adaptively adjusts the distribution according to the CPU load. Based on this, the robust CPU Clock Fingerprint Model based on the Adaptive Pareto Principle (CFMAP) is proposed. Via a KNN-based fingerprint recognition method, CFMAP solves the instability of existing CPU clock fingerprints under a high CPU load. Experiments show that the average recognition rate of CFMAP fingerprints is 96.82%. Moreover, they are highly robust against CPU load attacks and virtual machine attacks.

Recently, Logic Explained Networks (LENs) have been proposed as explainable-by-design neural models providing logic explanations for their predictions.However, these models have only been applied to vision and tabular data, and they mostly favour the generation of global explanations, while local ones tend to be noisy and verbose.For these reasons, we propose LEN p , improving local explanations by perturbing input words, and we test it on text classification.Our results show that (i) LEN p provides better local explanations than LIME in terms of sensitivity and faithfulness, and (ii) logic explanations are more useful and user-friendly than feature scoring provided by LIME as attested by a human survey.

Ensembles of deep learning models can be used for estimating predictive uncertainty. Existing ensemble approaches, however, introduce a high computational and memory cost limiting their applicability to real-time biosignal applications (e.g. ECG, EEG). To address these issues, we propose early exit ensembles (EEEs) for estimating predictive uncertainty via an implicit ensemble of early exits. In particular, EEEs are a collection of weight sharing sub-networks created by adding exit branches to any backbone neural network architecture. Empirical evaluation of EEEs demonstrates strong performance in accuracy and uncertainty metrics as well as computation gain highlighting the benefit of combining multiple structurally diverse models that can be jointly trained. Compared to state-of-the-art baselines (with an ensemble size of 5), EEEs can improve uncertainty metrics up to 2× while providing test-time speed-up and memory reduction of approx. 5×. Additionally, EEEs can improve accuracy up to 3.8 percentage points compared to single model baselines.

Automatically segmenting the anatomical structure of the heart from the cardiac magnetic resonance (CMR) images offers a great potential to augment the traditional healthcare strategy for the quantitative analysis of cardiac contractile function. Most of the existing CNN-based methods for cardiac segmentation tend to ignore the misalignment issues during the feature aggregation process and not fully use multi-scale context and contour information, which may lead to the unexpected misclassification caused by the falsely aligned contextual features and the discontinuity in the edge of segmentation maps. To resolve these issues, we proposed a context correlation aware network (CCA-Net). In CCA-Net, a volume correlation flow module was designed to align contour features and semantic features from adjacent levels, which offered the guidance to wrap low-resolution semantic features into high-resolution features. Besides, a residual gated squeeze module was utilized to explicitly model the boundaries and enhance the representations. Extensive experiments on the multi-sequence cardiac magnetic resonance segmentation challenge (MS-CMRSeg 2019) dataset and MICCAI challenge 2017 automatic cardiac diagnosis challenge (ACDC) dataset demonstrated that CCA-Net was superior to other state-of-the-art methods.

Polythetic classifications, based on shared patterns of features that need neither be universal nor constant among members of a class, are common in the natural world and greatly outnumber monothetic classifications over a set of features. We show that threshold meta-learners, such as Prototypical Networks, require an embedding dimension that is exponential in the number of task-relevant features to emulate these functions. In contrast, attentional classifiers, such as Matching Networks, are polythetic by default and able to solve these problems with a linear embedding dimension. However, we find that in the presence of task-irrelevant features, inherent to meta-learning problems, attentional models are susceptible to misclassification. To address this challenge, we propose a self-attention feature-selection mechanism that adaptively dilutes non-discriminative features. We demonstrate the effectiveness of our approach in meta-learning Boolean functions, and synthetic and real-world few-shot learning tasks.

Molecular property prediction is one of the fastest-growing applications of deep learning with critical real-world impacts. Although the 3D molecular graph structure is necessary for models to achieve strong performance on many tasks, it is infeasible to obtain 3D structures at the scale required by many real-world applications. To tackle this issue, we propose to use existing 3D molecular datasets to pre-train a model to reason about the geometry of molecules given only their 2D molecular graphs. Our method, called 3D Infomax, maximizes the mutual information between learned 3D summary vectors and the representations of a graph neural network (GNN). During fine-tuning on molecules with unknown geometry, the GNN is still able to produce implicit 3D information and uses it for downstream tasks. We show that 3D Infomax provides significant improvements for a wide range of properties, including a 22% average MAE reduction on QM9 quantum mechanical properties. Moreover, the learned representations can be effectively transferred between datasets in different molecular spaces.

This paper proposes an Android malware detection model based on Android Function Call Graph (FCG) and Denoising Graph Convolutional Neural Network. This study proposes a method to simplify the FCG to reduce its size, and a new method to construct vertex feature vectors. The model uses the subgraph network to detect the underlying structural features of the FCG and discover the confusion attack. A denoising graph neural network is applied to graph convolution to reduce the impact of obfuscation attacks.

Cellular sheaves equip graphs with a ``geometrical'' structure by assigning vector spaces and linear maps to nodes and edges. Graph Neural Networks (GNNs) implicitly assume a graph with a trivial underlying sheaf. This choice is reflected in the structure of the graph Laplacian operator, the properties of the associated diffusion equation, and the characteristics of the convolutional models that discretise this equation. In this paper, we use cellular sheaf theory to show that the underlying geometry of the graph is deeply linked with the performance of GNNs in heterophilic settings and their oversmoothing behaviour. By considering a hierarchy of increasingly general sheaves, we study how the ability of the sheaf diffusion process to achieve linear separation of the classes in the infinite time limit expands. At the same time, we prove that when the sheaf is non-trivial, discretised parametric diffusion processes have greater control than GNNs over their asymptotic behaviour. On the practical side, we study how sheaves can be learned from data. The resulting sheaf diffusion models have many desirable properties that address the limitations of classical graph diffusion equations (and corresponding GNN models) and obtain competitive results in heterophilic settings. Overall, our work provides new connections between GNNs and algebraic topology and would be of interest to both fields.

In the past few years, graph neural networks (GNNs) have become the de facto model of choice for graph classification. While, from the theoretical viewpoint, most GNNs can operate on graphs of any size, it is empirically observed that their classification performance degrades when they are applied on graphs with sizes that differ from those in the training data. Previous works have tried to tackle this issue in graph classification by providing the model with inductive biases derived from assumptions on the generative process of the graphs, or by requiring access to graphs from the test domain. The first strategy is tied to the quality of the assumptions made for the generative process, and requires the use of specific models designed after the explicit definition of the generative process of the data, leaving open the question of how to improve the performance of generic GNN models in general settings. On the other hand, the second strategy can be applied to any GNN, but requires access to information that is not always easy to obtain. In this work we consider the scenario in which we only have access to the training data, and we propose a regularization strategy that can be applied to any GNN to improve its generalization capabilities from smaller to larger graphs without requiring access to the test data. Our regularization is based on the idea of simulating a shift in the size of the training graphs using coarsening techniques, and enforcing the model to be robust to such a shift. Experimental results on standard datasets show that popular GNN models, trained on the 50% smallest graphs in the dataset and tested on the 10% largest graphs, obtain performance improvements of up to 30% when trained with our regularization strategy.

An effective aggregation of node features into a graph-level representation via readout functions is an essential step in numerous learning tasks involving graph neural networks. Typically, readouts are simple and non-adaptive functions designed such that the resulting hypothesis space is permutation invariant. Prior work on deep sets indicates that such readouts might require complex node embeddings that can be difficult to learn via standard neighborhood aggregation schemes. Motivated by this, we investigate the potential of adaptive readouts given by neural networks that do not necessarily give rise to permutation invariant hypothesis spaces. We argue that in some problems such as binding affinity prediction where molecules are typically presented in a canonical form it might be possible to relax the constraints on permutation invariance of the hypothesis space and learn a more effective model of the affinity by employing an adaptive readout function. Our empirical results demonstrate the effectiveness of neural readouts on more than 40 datasets spanning different domains and graph characteristics. Moreover, we observe a consistent improvement over standard readouts (i.e., sum, max, and mean) relative to the number of neighborhood aggregation iterations and different convolutional operators.

In many real world problems, features do not act alone but in combination with each other. For example, in genomics, diseases might not be caused by any single mutation but require the presence of multiple mutations. Prior work on feature selection either seeks to identify individual features or can only determine relevant groups from a predefined set. We investigate the problem of discovering groups of predictive features without predefined grouping. To do so, we define predictive groups in terms of linear and non-linear interactions between features. We introduce a novel deep learning architecture that uses an ensemble of feature selection models to find predictive groups, without requiring candidate groups to be provided. The selected groups are sparse and exhibit minimum overlap. Furthermore, we propose a new metric to measure similarity between discovered groups and the ground truth. We demonstrate the utility our model on multiple synthetic tasks and semi-synthetic chemistry datasets, where the ground truth structure is known, as well as an image dataset and a real-world cancer dataset.

Geometric deep learning has broad applications in biology, a domain where relational structure in data is often intrinsic to modelling the underlying phenomena. Currently, efforts in both geometric deep learning and, more broadly, deep learning applied to biomolecular tasks have been hampered by a scarcity of appropriate datasets accessible to domain specialists and machine learning researchers alike. To address this, we introduce Graphein as a turn-key tool for transforming raw data from widely-used bioinformatics databases into machine learning-ready datasets in a high-throughput and flexible manner. Graphein is a Python library for constructing graph and surface-mesh representations of biomolecular structures, such as proteins, nucleic acids and small molecules, and biological interaction networks for computational analysis and machine learning. Graphein provides utilities for data retrieval from widely-used bioinformatics databases for structural data, including the Protein Data Bank, the AlphaFold Structure Database, chemical data from ZINC and ChEMBL, and for biomolecular interaction networks from STRINGdb, BioGrid, TRRUST and RegNetwork. The library interfaces with popular geometric deep learning libraries: DGL, Jraph, PyTorch Geometric and PyTorch3D though remains framework agnostic as it is built on top of the PyData ecosystem to enable inter-operability with scientific computing tools and libraries. Graphein is designed to be highly flexible, allowing the user to specify each step of the data preparation, scalable to facilitate working with large protein complexes and interaction graphs, and contains useful pre-processing tools for preparing experimental files. Graphein facilitates network-based, graph-theoretic and topological analyses of structural and interaction datasets in a high-throughput manner. We envision that Graphein will facilitate developments in computational biology, graph representation learning and drug discovery. Availability and implementation: Graphein is written in Python. Source code, example usage and tutorials, datasets, and documentation are made freely available under the MIT License at the following URL: https://anonymous.4open.science/r/graphein-3472/README.md

Deploying AI-powered systems requires trustworthy models supporting effective human interactions, going beyond raw prediction accuracy. Concept bottleneck models promote trustworthiness by conditioning classification tasks on an intermediate level of human-like concepts. This enables human interventions which can correct mispredicted concepts to improve the model's performance. However, existing concept bottleneck models are unable to find optimal compromises between high task accuracy, robust concept-based explanations, and effective interventions on concepts---particularly in real-world conditions where complete and accurate concept supervisions are scarce. To address this, we propose Concept Embedding Models, a novel family of concept bottleneck models which goes beyond the current accuracy-vs-interpretability trade-off by learning interpretable high-dimensional concept representations. Our experiments demonstrate that Concept Embedding Models (1) attain better or competitive task accuracy w.r.t. standard neural models without concepts, (2) provide concept representations capturing meaningful semantics including and beyond their ground truth labels, (3) support test-time concept interventions whose effect in test accuracy surpasses that in standard concept bottleneck models, and (4) scale to real-world conditions where complete concept supervisions are scarce.

MOTIVATION: Single-cell RNA sequencing allows high-resolution views of individual cells for libraries of up to millions of samples, thus motivating the use of deep learning for analysis. In this study, we introduce the use of graph neural networks for the unsupervised exploration of scRNA-seq data by developing a variational graph autoencoder architecture with graph attention layers that operates directly on the connectivity between cells, focusing on dimensionality reduction and clustering. With the help of several case studies, we show that our model, named CellVGAE, can be effectively used for exploratory analysis even on challenging datasets, by extracting meaningful features from the data and providing the means to visualize and interpret different aspects of the model. RESULTS: We show that CellVGAE is more interpretable than existing scRNA-seq variational architectures by analysing the graph attention coefficients. By drawing parallels with other scRNA-seq studies on interpretability, we assess the validity of the relationships modelled by attention, and furthermore, we show that CellVGAE can intrinsically capture information such as pseudotime and NF-ĸB activation dynamics, the latter being a property that is not generally shared by existing neural alternatives. We then evaluate the dimensionality reduction and clustering performance on 9 difficult and well-annotated datasets by comparing with three leading neural and non-neural techniques, concluding that CellVGAE outperforms competing methods. Finally, we report a decrease in training times of up to × 20 on a dataset of 1.3 million cells compared to existing deep learning architectures. AVAILABILITYAND IMPLEMENTATION: The CellVGAE code is available at https://github.com/davidbuterez/CellVGAE. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Gene expression data are commonly used at the intersection of cancer research and machine learning for better understanding of the molecular status of tumour tissue. Deep learning predictive models have been employed for gene expression data due to their ability to scale and remove the need for manual feature engineering. However, gene expression data are often very high dimensional, noisy and presented with a low number of samples. This poses significant problems for learning algorithms: models often overfit, learn noise and struggle to capture biologically relevant information. In this article, we utilize external biological knowledge embedded within structures of gene interaction graphs such as protein-protein interaction (PPI) networks to guide the construction of predictive models. RESULTS: We present Gene Interaction Network Constrained Construction (GINCCo), an unsupervised method for automated construction of computational graph models for gene expression data that are structurally constrained by prior knowledge of gene interaction networks. We employ this methodology in a case study on incorporating a PPI network in cancer phenotype prediction tasks. Our computational graphs are structurally constructed using topological clustering algorithms on the PPI networks which incorporate inductive biases stemming from network biology research on protein complex discovery. Each of the entities in the GINCCo computational graph represents biological entities such as genes, candidate protein complexes and phenotypes instead of arbitrary hidden nodes of a neural network. This provides a biologically relevant mechanism for model regularization yielding strong predictive performance while drastically reducing the number of model parameters and enabling guided post-hoc enrichment analyses of influential gene sets with respect to target phenotypes. Our experiments analysing a variety of cancer phenotypes show that GINCCo often outperforms support vector machine, Fully Connected Multi-layer Perceptrons (MLP) and Randomly Connected MLPs despite greatly reduced model complexity. AVAILABILITY AND IMPLEMENTATION: https://github.com/paulmorio/gincco contains the source code for our approach. We also release a library with algorithms for protein complex discovery within PPI networks at https://github.com/paulmorio/protclus. This repository contains implementations of the clustering algorithms used in this article. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: High-throughput gene expression can be used to address a wide range of fundamental biological problems, but datasets of an appropriate size are often unavailable. Moreover, existing transcriptomics simulators have been criticized because they fail to emulate key properties of gene expression data. In this article, we develop a method based on a conditional generative adversarial network to generate realistic transcriptomics data for Escherichia coli and humans. We assess the performance of our approach across several tissues and cancer-types. RESULTS: We show that our model preserves several gene expression properties significantly better than widely used simulators, such as SynTReN or GeneNetWeaver. The synthetic data preserve tissue- and cancer-specific properties of transcriptomics data. Moreover, it exhibits real gene clusters and ontologies both at local and global scales, suggesting that the model learns to approximate the gene expression manifold in a biologically meaningful way. AVAILABILITY AND IMPLEMENTATION: Code is available at: https://github.com/rvinas/adversarial-gene-expression. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The computing power of various Internet of Things (IoT) devices is quite different. To enable IoT devices with lower computing power to perform machine learning, all nodes can only train smaller models, which results in the waste of computing power for high-performance devices. In this article, a heterogeneous model fusion federated learning (HFL) mechanism is proposed. Each node trains learning models of different scales according to its own computing capabilities. After receiving the gradient trained by each node, the parameter server (PS) corrects the received gradient with the repeat matrix, and then update the corresponding region of the global model according to the mapping matrix. After all update operations are over, the PS assigns the compressed model to the corresponding node. This article uses a variety of experimental schemes to evaluate the proposed method, including three data sets, two model structures, and three computational complexity levels. The proposed method have been proved it not only maximizes the use of the unbalanced computing power of edge nodes, but also enables different structural models to compensate for the shortcomings of others, improving the overall performance.

Resting-state functional magnetic resonance imaging (rs-fMRI) has been successfully employed to understand the organisation of the human brain. Typically, the brain is parcellated into regions of interest (ROIs) and modelled as a graph where each ROI represents a node and association measures between ROI-specific blood-oxygen-level-dependent (BOLD) time series are edges. Recently, graph neural networks (GNNs) have seen a surge in popularity due to their success in modelling unstructured relational data. The latest developments with GNNs, however, have not yet been fully exploited for the analysis of rs-fMRI data, particularly with regards to its spatio-temporal dynamics. In this paper, we present a novel deep neural network architecture which combines both GNNs and temporal convolutional networks (TCNs) in order to learn from both the spatial and temporal components of rs-fMRI data in an end-to-end fashion. In particular, this corresponds to intra-feature learning (i.e., learning temporal dynamics with TCNs) as well as inter-feature learning (i.e., leveraging interactions between ROI-wise dynamics with GNNs). We evaluate our model with an ablation study using 35,159 samples from the UK Biobank rs-fMRI database, as well as in the smaller Human Connectome Project (HCP) dataset, both in a unimodal and in a multimodal fashion. We also demonstrate that out architecture contains explainability-related features which easily map to realistic neurobiological insights. We suggest that this model could lay the groundwork for future deep learning architectures focused on leveraging the inherently and inextricably spatio-temporal nature of rs-fMRI data.

Over the past decade, deep learning has had a revolutionary impact on a broad range of fields such as computer vision and image processing, computational photography, medical imaging and speech and language analysis and synthesis etc. Deep learning technologies are estimated to have added billions in business value, created new markets, and transformed entire industrial segments. Most of today’s successful deep learning methods such as Convolutional Neural Networks (CNNs) rely on classical signal processing models that limit their applicability to data with underlying Euclidean grid-like structure, e.g., images or acoustic signals. Yet, many applications deal with non-Euclidean (graph- or manifold-structured) data. For example, in social network analysis the users and their attributes are generally modeled as signals on the vertices of graphs. In biology protein-to-protein interactions are modeled as graphs. In computer vision & graphics 3D objects are modeled as meshes or point clouds. Furthermore, a graph representation is a very natural way to describe interactions between objects or signals. The classical deep learning paradigm on Euclidean domains falls short in providing appropriate tools for such kind of data. Until recently, the lack of deep learning models capable of correctly dealing with non-Euclidean data has been a major obstacle in these fields. This special section addresses the need to bring together leading efforts in non-Euclidean deep learning across all communities. From the papers that the special received twelve were selected for publication. The selected papers can naturally fall in three distinct categories: (a) methodologies that advance machine learning on data that are represented as graphs, (b) methodologies that advance machine learning on manifold-valued data, and (c) applications of machine learning methodologies on non-Euclidean spaces in computer vision and medical imaging. We briefly review the accepted papers in each of the groups.

Compressed sensing (CS) has been playing a key role in accelerating the magnetic resonance imaging (MRI) acquisition process. With the resurgence of artificial intelligence, deep neural networks and CS algorithms are being integrated to redefine the state of the art of fast MRI. The past several years have witnessed substantial growth in the complexity, diversity, and performance of deep-learning-based CS techniques that are dedicated to fast MRI. In this meta-analysis, we systematically review the deep-learning-based CS techniques for fast MRI, describe key model designs, highlight breakthroughs, and discuss promising directions. We have also introduced a comprehensive analysis framework and a classification system to assess the pivotal role of deep learning in CS-based acceleration for MRI.

The pairwise interaction paradigm of graph machine learning has predominantly governed the modelling of relational systems. However, graphs alone cannot capture the multi-level interactions present in many complex systems and the expressive power of such schemes was proven to be limited. To overcome these limitations, we propose Message Passing Simplicial Networks (MPSNs), a class of models that perform message passing on simplicial complexes (SCs). To theoretically analyse the expressivity of our model we introduce a Simplicial Weisfeiler-Lehman (SWL) colouring procedure for distinguishing non-isomorphic SCs. We relate the power of SWL to the problem of distinguishing non-isomorphic graphs and show that SWL and MPSNs are strictly more powerful than the WL test and not less powerful than the 3-WL test. We deepen the analysis by comparing our model with traditional graph neural networks (GNNs) with ReLU activations in terms of the number of linear regions of the functions they can represent. We empirically support our theoretical claims by showing that MPSNs can distinguish challenging strongly regular graphs for which GNNs fail and, when equipped with orientation equivariant layers, they can improve classification accuracy in oriented SCs compared to a GNN baseline.

This paper presents a new approach for assembling graph neural networks based on framelet transforms. The latter provides a multi-scale representation for graph-structured data. We decompose an input graph into low-pass and high-pass frequencies coefficients for network training, which then defines a framelet-based graph convolution. The framelet decomposition naturally induces a graph pooling strategy by aggregating the graph feature into low-pass and high-pass spectra, which considers both the feature values and geometry of the graph data and conserves the total information. The graph neural networks with the proposed framelet convolution and pooling achieve state-of-the-art performance in many node and graph prediction tasks. Moreover, we propose shrinkage as a new activation for the framelet convolution, which thresholds high-frequency information at different scales. Compared to ReLU, shrinkage activation improves model performance on denoising and signal compression: noises in both node and structure can be significantly reduced by accurately cutting off the high-pass coefficients from framelet decomposition, and the signal can be compressed to less than half its original size with well-preserved prediction performance.

Graph Neural Networks (GNNs) are limited in their expressive power, struggle with long-range interactions and lack a principled way to model higher-order structures. These problems can be attributed to the strong coupling between the computational graph and the input graph structure. The recently proposed Message Passing Simplicial Networks naturally decouple these elements by performing message passing on the clique complex of the graph. Nevertheless, these models can be severely constrained by the rigid combinatorial structure of Simplicial Complexes (SCs). In this work, we extend recent theoretical results on SCs to regular Cell Complexes, topological objects that flexibly subsume SCs and graphs. We show that this generalisation provides a powerful set of graph "lifting" transformations, each leading to a unique hierarchical message passing procedure. The resulting methods, which we collectively call CW Networks (CWNs), are strictly more powerful than the WL test and not less powerful than the 3-WL test. In particular, we demonstrate the effectiveness of one such scheme, based on rings, when applied to molecular graph problems. The proposed architecture benefits from provably larger expressivity than commonly used GNNs, principled modelling of higher-order signals and from compressing the distances between nodes. We demonstrate that our model achieves state-of-the-art results on a variety of molecular datasets.

The development of data-dependent heuristics and representations for biological sequences that reflect their evolutionary distance is critical for large-scale biological research. However, popular machine learning approaches, based on continuous Euclidean spaces, have struggled with the discrete combinatorial formulation of the edit distance that models evolution and the hierarchical relationship that characterises real-world datasets. We present Neural Distance Embeddings (NeuroSEED), a general framework to embed sequences in geometric vector spaces, and illustrate the effectiveness of the hyperbolic space that captures the hierarchical structure and provides an average 38% reduction in embedding RMSE against the best competing geometry. The capacity of the framework and the significance of these improvements are then demonstrated devising supervised and unsupervised NeuroSEED approaches to multiple core tasks in bioinformatics. Benchmarked with common baselines, the proposed approaches display significant accuracy and/or runtime improvements on real-world datasets. As an example for hierarchical clustering, the proposed pretrained and from-scratch methods match the quality of competing baselines with 30x and 15x runtime reduction, respectively.

Although chemotherapy is the first-line treatment for ovarian cancer (OCa) patients, chemoresistance (CR) decreases their progression-free survival. This paper investigates the genetic interaction (GI) related to OCa-CR. To decrease the complexity of establishing gene networks, individual signature genes related to OCa-CR are identified using a gradient boosting decision tree algorithm. Additionally, the genetic interaction coefficient (GIC) is proposed to measure the correlation of two signature genes quantitatively and explain their joint influence on OCa-CR. Gene pair that possesses high GIC is identified as signature pair. A total of 24 signature gene pairs are selected that include 10 individual signature genes and the influence of signature gene pairs on OCa-CR is explored. Finally, a signature gene pair-based prediction of OCa-CR is identified. The area under curve (AUC) is a widely used performance measure for machine learning prediction. The AUC of signature gene pair reaches 0.9658, whereas the AUC of individual signature gene-based prediction is 0.6823 only. The identified signature gene pairs not only build an efficient GI network of OCa-CR but also provide an interesting way for OCa-CR prediction. This improvement shows that our proposed method is a useful tool to investigate GI related to OCa-CR.

The novel coronavirus (2019-nCoV) has recently emerged, causing COVID-19 outbreaks and significant societal/global disruption. Importantly, COVID-19 infection resembles SARS-like complications. However, the lack of knowledge about the underlying genetic mechanisms of COVID-19 warrants the development of prospective control measures. In this study, we employed whole-genome alignment and digital DNA-DNA hybridization analyses to assess genomic linkage between 2019-nCoV and other coronaviruses. To understand the pathogenetic behavior of 2019-nCoV, we compared gene expression datasets of viral infections closest to 2019-nCoV with four COVID-19 clinical presentations followed by functional enrichment of shared dysregulated genes. Potential chemical antagonists were also identified using protein-chemical interaction analysis. Based on phylogram analysis, the 2019-nCoV was found genetically closest to SARS-CoVs. In addition, we identified 562 upregulated and 738 downregulated genes (adj. P ≤ 0.05) with SARS-CoV infection. Among the dysregulated genes, SARS-CoV shared ≤19 upregulated and ≤22 downregulated genes with each of different COVID-19 complications. Notably, upregulation of BCL6 and PFKFB3 genes was common to SARS-CoV, pneumonia and severe acute respiratory syndrome, while they shared CRIP2, NSG1 and TNFRSF21 genes in downregulation. Besides, 14 genes were common to different SARS-CoV comorbidities that might influence COVID-19 disease. We also observed similarities in pathways that can lead to COVID-19 and SARS-CoV diseases. Finally, protein-chemical interactions suggest cyclosporine, resveratrol and quercetin as promising drug candidates against COVID-19 as well as other SARS-like viral infections. The pathogenetic analyses, along with identified biomarkers, signaling pathways and chemical antagonists, could prove useful for novel drug development in the fight against the current global 2019-nCoV pandemic.

MOTIVATION: One of the branches of Systems Biology is focused on a deep understanding of underlying regulatory networks through the analysis of the biomolecules oscillations and their interplay. Synthetic Biology exploits gene or/and protein regulatory networks towards the design of oscillatory networks for producing useful compounds. Therefore, at different levels of application and for different purposes, the study of biomolecular oscillations can lead to different clues about the mechanisms underlying living cells. It is known that network-level interactions involve more than one type of biomolecule as well as biological processes operating at multiple omic levels. Combining network/pathway-level information with genetic information it is possible to describe well-understood or unknown bacterial mechanisms and organism-specific dynamics. RESULTS: Following the methodologies used in signal processing and communication engineering, a methodology is introduced to identify and quantify the extent of multi-omic oscillations. These are due to the process of multi-omic integration and depend on the gene positions on the chromosome. Ad hoc signal metrics are designed to allow further biotechnological explanations and provide important clues about the oscillatory nature of the pathways and their regulatory circuits. Our algorithms designed for the analysis of multi-omic signals are tested and validated on 11 different bacteria for thousands of multi-omic signals perturbed at the network level by different experimental conditions. Information on the order of genes, codon usage, gene expression and protein molecular weight is integrated at three different functional levels. Oscillations show interesting evidence that network-level multi-omic signals present a synchronized response to perturbations and evolutionary relations along taxa. AVAILABILITY AND IMPLEMENTATION: The algorithms, the code (in language R), the tool, the pipeline and the whole dataset of multi-omic signal metrics are available at: https://github.com/lodeguns/Multi-omicSignals. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: High-throughput sequencing Chromosome Conformation Capture (Hi-C) allows the study of DNA interactions and 3D chromosome folding at the genome-wide scale. Usually, these data are represented as matrices describing the binary contacts among the different chromosome regions. On the other hand, a graph-based representation can be advantageous to describe the complex topology achieved by the DNA in the nucleus of eukaryotic cells. METHODS: Here we discuss the use of a graph database for storing and analysing data achieved by performing Hi-C experiments. The main issue is the size of the produced data and, working with a graph-based representation, the consequent necessity of adequately managing a large number of edges (contacts) connecting nodes (genes), which represents the sources of information. For this, currently available graph visualisation tools and libraries fall short with Hi-C data. The use of graph databases, instead, supports both the analysis and the visualisation of the spatial pattern present in Hi-C data, in particular for comparing different experiments or for re-mapping omics data in a space-aware context efficiently. In particular, the possibility of describing graphs through statistical indicators and, even more, the capability of correlating them through statistical distributions allows highlighting similarities and differences among different Hi-C experiments, in different cell conditions or different cell types. RESULTS: These concepts have been implemented in NeoHiC, an open-source and user-friendly web application for the progressive visualisation and analysis of Hi-C networks based on the use of the Neo4j graph database (version 3.5). CONCLUSION: With the accumulation of more experiments, the tool will provide invaluable support to compare neighbours of genes across experiments and conditions, helping in highlighting changes in functional domains and identifying new co-organised genomic compartments.

BACKGROUND: Single-cell RNA sequencing (scRNA-Seq) experiments are gaining ground to study the molecular processes that drive normal development as well as the onset of different pathologies. Finding an effective and efficient low-dimensional representation of the data is one of the most important steps in the downstream analysis of scRNA-Seq data, as it could provide a better identification of known or putatively novel cell-types. Another step that still poses a challenge is the integration of different scRNA-Seq datasets. Though standard computational pipelines to gain knowledge from scRNA-Seq data exist, a further improvement could be achieved by means of machine learning approaches. RESULTS: Autoencoders (AEs) have been effectively used to capture the non-linearities among gene interactions of scRNA-Seq data, so that the deployment of AE-based tools might represent the way forward in this context. We introduce here scAEspy, a unifying tool that embodies: (1) four of the most advanced AEs, (2) two novel AEs that we developed on purpose, (3) different loss functions. We show that scAEspy can be coupled with various batch-effect removal tools to integrate data by different scRNA-Seq platforms, in order to better identify the cell-types. We benchmarked scAEspy against the most used batch-effect removal tools, showing that our AE-based strategies outperform the existing solutions. CONCLUSIONS: scAEspy is a user-friendly tool that enables using the most recent and promising AEs to analyse scRNA-Seq data by only setting up two user-defined parameters. Thanks to its modularity, scAEspy can be easily extended to accommodate new AEs to further improve the downstream analysis of scRNA-Seq data. Considering the relevant results we achieved, scAEspy can be considered as a starting point to build a more comprehensive toolkit designed to integrate multi single-cell omics.

Single image super-resolution (SISR) aims to obtain a high-resolution output from one low-resolution image. Currently, deep learning-based SISR approaches have been widely discussed in medical image processing, because of their potential to achieve high-quality, high spatial resolution images without the cost of additional scans. However, most existing methods are designed for scale-specific SR tasks and are unable to generalize over magnification scales. In this paper, we propose an approach for medical image arbitrary-scale super-resolution (MIASSR), in which we couple meta-learning with generative adversarial networks (GANs) to super-resolve medical images at any scale of magnification in [Formula: see text]. Compared to state-of-the-art SISR algorithms on single-modal magnetic resonance (MR) brain images (OASIS-brains) and multi-modal MR brain images (BraTS), MIASSR achieves comparable fidelity performance and the best perceptual quality with the smallest model size. We also employ transfer learning to enable MIASSR to tackle SR tasks of new medical modalities, such as cardiac MR images (ACDC) and chest computed tomography images (COVID-CT). The source code of our work is also public. Thus, MIASSR has the potential to become a new foundational pre-/post-processing step in clinical image analysis tasks such as reconstruction, image quality enhancement, and segmentation.

The COVID-19 global pandemic is a threat not only to the health of millions of individuals, but also to the stability of infrastructure and economies around the world. The disease will inevitably place an overwhelming burden on healthcare systems that cannot be effectively dealt with by existing facilities or responses based on conventional approaches. We believe that a rigorous clinical and societal response can only be mounted by using intelligence derived from a variety of data sources to better utilize scarce healthcare resources, provide personalized patient management plans, inform policy, and expedite clinical trials. In this paper, we introduce five of the most important challenges in responding to COVID-19 and show how each of them can be addressed by recent developments in machine learning (ML) and artificial intelligence (AI). We argue that the integration of these techniques into local, national, and international healthcare systems will save lives, and propose specific methods by which implementation can happen swiftly and efficiently. We offer to extend these resources and knowledge to assist policymakers seeking to implement these techniques.

IoT technology has been widely valued and applied, and the resulting massive IoT data brings many challenges to the traditional centralized data management, such as performance, privacy, and security challenges. This paper proposes an IoT data access control scheme that combines attribute-based encryption (ABE) and blockchain technology. Symmetric encryption and ABE algorithms are utilized to realize fine-grained access control and ensure the security and openness of IoT data. Moreover, blockchain technology is combined with distributed storage to solve the storage bottleneck of blockchain systems. Only the hash values of the data, the hash values of the ciphertext location, the access control policy, and other important information are stored on the blockchain. In this scheme, smart contract is used to implement access control. The results of experiments demonstrate that the proposed scheme can effectively protect the security and privacy of IoT data and realize the secure sharing of data.

Reinforcement Learning (RL) has achieved impressive performance in many complex environments due to the integration with Deep Neural Networks (DNNs). At the same time, Genetic Algorithms (GAs), often seen as a competing approach to RL, had limited success in scaling up to the DNNs required to solve challenging tasks. Contrary to this dichotomic view, in the physical world, evolution and learning are complementary processes that continuously interact. The recently proposed Evolutionary Reinforcement Learning (ERL) framework has demonstrated mutual benefits to performance when combining the two methods. However, ERL has not fully addressed the scalability problem of GAs. In this paper, we show that this problem is rooted in an unfortunate combination of a simple genetic encoding for DNNs and the use of traditional biologically-inspired variation operators. When applied to these encodings, the standard operators are destructive and cause catastrophic forgetting of the traits the networks acquired. We propose a novel algorithm called Proximal Distilled Evolutionary Reinforcement Learning (PDERL) that is characterised by a hierarchical integration between evolution and learning. The main innovation of PDERL is the use of learning-based variation operators that compensate for the simplicity of the genetic representation. Unlike traditional operators, our proposals meet the functional requirements of variation operators when applied on directly-encoded DNNs. We evaluate PDERL in five robot locomotion settings from the OpenAI gym. Our method outperforms ERL, as well as two state-of-the-art RL algorithms, PPO and TD3, in all tested environments.

Multi-Agent Reinforcement Learning (MARL) encompasses a powerful class of methodologies that have been applied in a wide range of fields. An effective way to further empower these methodologies is to develop approaches and tools that could expand their interpretability and explainability. In this work, we introduce MARLeME: a MARL model extraction library, designed to improve explainability of MARL systems by approximating them with symbolic models. Symbolic models offer a high degree of interpretability, well-defined properties, and verifiable behaviour. Consequently, they can be used to inspect and better understand the underlying MARL systems and corresponding MARL agents, as well as to replace all/some of the agents that are particularly safety and security critical. In this work, we demonstrate how MARLeME can be applied to two well-known case studies (Cooperative Navigation and RoboCup Takeaway), using extracted models based on Abstract Argumentation.

Graph Neural Networks (GNNs) have been shown to be effective models for different predictive tasks on graph-structured data. Recent work on their expressive power has focused on isomorphism tasks and countable feature spaces. We extend this theoretical framework to include continuous features---which occur regularly in real-world input domains and within the hidden layers of GNNs---and we demonstrate the requirement for multiple aggregation functions in this context. Accordingly, we propose Principal Neighbourhood Aggregation (PNA), a novel architecture combining multiple aggregators with degree-scalers (which generalize the sum aggregator). Finally, we compare the capacity of different models to capture and exploit the graph structure via a novel benchmark containing multiple tasks taken from classical graph theory, alongside existing benchmarks from real-world domains, all of which demonstrate the strength of our model. With this work we hope to steer some of the GNN research towards new aggregation methods which we believe are essential in the search for powerful and robust models.

We examine the problem of controlling divergences for latent space regularisation in variational autoencoders. Specifically, when aiming to reconstruct example $x\in\mathbb{R}^{m}$ via latent space $z\in\mathbb{R}^{n}$ ($n\leq m$), while balancing this against the need for generalisable latent representations. We present a regularisation mechanism based on the {\em skew-geometric Jensen-Shannon divergence} $\left(\textrm{JS}^{\textrm{G}_{\alpha}}\right)$. We find a variation in $\textrm{JS}^{\textrm{G}_{\alpha}}$, motivated by limiting cases, which leads to an intuitive interpolation between forward and reverse KL in the space of both distributions and divergences. We motivate its potential benefits for VAEs through low-dimensional examples, before presenting quantitative and qualitative results. Our experiments demonstrate that skewing our variant of $\textrm{JS}^{\textrm{G}_{\alpha}}$, in the context of $\textrm{JS}^{\textrm{G}_{\alpha}}$-VAEs, leads to better reconstruction and generation when compared to several baseline VAEs. Our approach is entirely unsupervised and utilises only one hyperparameter which can be easily interpreted in latent space.

Graph neural networks (GNNs) extends the functionality of traditional neural networks to graph-structured data. Similar to CNNs, an optimized design of graph convolution and pooling is key to success. Borrowing ideas from physics, we propose a path integral based graph neural networks (PAN) for classification and regression tasks on graphs. Specifically, we consider a convolution operation that involves every path linking the message sender and receiver with learnable weights depending on the path length, which corresponds to the maximal entropy random walk. It generalizes the graph Laplacian to a new transition matrix we call \emph{maximal entropy transition} (MET) matrix derived from a path integral formalism. Importantly, the diagonal entries of the MET matrix are directly related to the subgraph centrality, thus lead to a natural and adaptive pooling mechanism. PAN provides a versatile framework that can be tailored for different graph data with varying sizes and structures. We can view most existing GNN architectures as special cases of PAN. Experimental results show that PAN achieves state-of-the-art performance on various graph classification/regression tasks, including a new benchmark dataset from statistical mechanics we propose to boost applications of GNN in physical sciences.

Neural Ordinary Differential Equations (NODEs) are a new class of models that transform data continuously through infinite-depth architectures. The continuous nature of NODEs has made them particularly suitable for learning the dynamics of complex physical systems. While previous work has mostly been focused on first order ODEs, the dynamics of many systems, especially in classical physics, are governed by second order laws. In this work, we consider Second Order Neural ODEs (SONODEs). We show how the adjoint sensitivity method can be extended to SONODEs and prove that the optimisation of a first order coupled ODE is equivalent and computationally more efficient. Furthermore, we extend the theoretical understanding of the broader class of Augmented NODEs (ANODEs) by showing they can also learn higher order dynamics with a minimal number of augmented dimensions, but at the cost of interpretability. This indicates that the advantages of ANODEs go beyond the extra space offered by the augmented dimensions, as originally thought. Finally, we compare SONODEs and ANODEs on synthetic and real dynamical systems and demonstrate that the inductive biases of the former generally result in faster training and better performance.

A fundamental problem of interest to policy makers, urban planners, and other stakeholders involved in urban development is assessing the impact of planning and construction activities on mobility flows. This is a challenging task due to the different spatial, temporal, social, and economic factors influencing urban mobility flows. These flows, along with the influencing factors, can be modelled as attributed graphs with both node and edge features characterising locations in a city and the various types of relationships between them. In this paper, we address the problem of assessing origin-destination (OD) car flows between a location of interest and every other location in a city, given their features and the structural characteristics of the graph. We propose three neural network architectures, including graph neural networks (GNN), and conduct a systematic comparison between the proposed methods and state-of-the-art spatial interaction models, their modifications, and machine learning approaches. The objective of the paper is to address the practical problem of estimating potential flow between an urban project location and other locations in the city, where the features of the project location are known in advance. We evaluate the performance of the models on a regression task using a custom data set of attributed car OD flows in London. We also visualise the model performance by showing the spatial distribution of flow residuals across London.

Coeliac disease (CD) is a complex, multifactorial pathology caused by different factors, such as nutrition, immunological response and genetic factors. Many autoimmune diseases are comorbidities for CD, and a comprehensive and integrated analysis with bioinformatics approaches can help in evaluating the interconnections among all the selected pathologies. We first performed a detailed survey of gene expression data available in public repositories on CD and less commonly considered comorbidities. Then we developed an innovative pipeline that integrates gene expression, cell-type data and online resources (e.g. a list of comorbidities from the literature), using bioinformatics methods such as gene set enrichment analysis and semantic similarity. Our pipeline is written in R language, available at the following link: http://bioinformatica.isa.cnr.it/COELIAC_DISEASE/SCRIPTS/. We found a list of common differential expressed genes, gene ontology terms and pathways among CD and comorbidities and the closeness among the selected pathologies by means of disease ontology terms. Physicians and other researchers, such as molecular biologists, systems biologists and pharmacologists can use it to analyze pathology in detail, from differential expressed genes to ontologies, performing a comparison with the pathology comorbidities or with other diseases.

Personal keystroke modes are difficult to imitate and can therefore be used for identity authentication. The keystroke habits of a person can be learned according to the keystroke data generated when the person inputs free text. Detecting a user's keystroke habits as the user enters text can continuously verify the user's identity without affecting user input. The method proposed in this paper authenticates users via their keystrokes when they type free text. The user keystroke data is divided into a fixed-length keystroke sequence, which is then converted into a keystroke vector sequence according to the time feature of the keystroke. A model that combines a convolutional neural network and a recursive neural network is used to learn a sequence of individual keystroke vectors to obtain individual keystroke features for identity authentication. The model is tested using two open datasets, and the best false rejection rate (FRR) is found to be (2.07%,6.61%), the best false acceptance rate (FAR) is found to be (3.26%, 5.31%), and the best equal error rate (EER) is found to be (2.67%, 5.97%).

Abstract This editorial introduces the articles selected for the special issue concerning the International Conferences on Parallel, Distributed, and Network‐Based Processing, which provided insights related to the efficient exploitation of parallel and distributed architectures, including power‐aware computing, application scheduling, and application development for GPUs.

Computational Biology is a fast-growing field that is enriched by different data-driven methodological approaches and by findings and applications in a broad range of biological areas. Fundamental to these approaches are the mathematical and computational models used to describe the different state s at microscopic (for example a biochemical reaction), mesoscopic (the signalling effects at tissue level), and macroscopic levels (physiological and pathological effects) of biological processes. In this paper we address the problem of combining two powerful classes of methodologies: Flux Balance Analysis (FBA) methods which are now producing a revolution in biotechnology and medicine, and Petri Nets (PNs) which allow system generalisation and are central to various mathematical treatments, for example Ordinary Differential Equation (ODE) specification of the biosystem under study. While the former is limited to modelling metabolic networks, i.e. does not account for intermittent dynamical signalling events, the latter is hampered by the need for a large amount of metabolic data. A first result presented in this paper is the identification of three types of cross-talks between PNs and FBA methods and their dependencies on available data. We exemplify our insights with the analysis of a pancreatic cancer model. We discuss how our reasoning framework provides a biologically and mathematically grounded decision making setting for the integration of regulatory, signalling, and metabolic networks and greatly increases model interpretability and reusability. We discuss how the parameters of PN and FBA models can be tuned and combined together so to highlight the computational effort needed to perform this task. We conclude with speculations and suggestions on this new promising research direction.

In recent years, there have been numerous developments toward solving multimodal tasks, aiming to learn a stronger representation than through a single modality. Certain aspects of the data can be particularly useful in this case-for example, correlations in the space or time domain across modalities-but should be wisely exploited in order to benefit from their full predictive potential. We propose two deep learning architectures with multimodal cross connections that allow for dataflow between several feature extractors (XFlow). Our models derive more interpretable features and achieve better performances than models that do not exchange representations, usefully exploiting correlations between audio and visual data, which have a different dimensionality and are nontrivially exchangeable. This article improves on the existing multimodal deep learning algorithms in two essential ways: 1) it presents a novel method for performing cross modality (before features are learned from individual modalities) and 2) extends the previously proposed cross connections that only transfer information between the streams that process compatible data. Illustrating some of the representations learned by the connections, we analyze their contribution to the increase in discrimination ability and reveal their compatibility with a lip-reading network intermediate representation. We provide the research community with Digits, a new data set consisting of three data types extracted from videos of people saying the digits 0-9. Results show that both cross-modal architectures outperform their baselines (by up to 11.5%) when evaluated on the AVletters, CUAVE, and Digits data sets, achieving the state-of-the-art results.

Word embedding techniques heavily rely on the abundance of training data for individual words. Given the Zipfian distribution of words in natural language texts, a large number of words do not usually appear frequently or at all in the training data. In this paper we put forward a technique that exploits the knowledge encoded in lexical resources, such as WordNet, to induce embeddings for unseen words. Our approach adapts graph embedding and cross-lingual vector space transformation techniques in order to merge lexical knowledge encoded in ontologies with that derived from corpus statistics. We show that the approach can provide consistent performance improvements across multiple evaluation benchmarks: in-vitro, on multiple rare word similarity datasets, and invivo, in two downstream text classification tasks.

BACKGROUND: In order to fully characterize the genome of an individual, the reconstruction of the two distinct copies of each chromosome, called haplotypes, is essential. The computational problem of inferring the full haplotype of a cell starting from read sequencing data is known as haplotype assembly, and consists in assigning all heterozygous Single Nucleotide Polymorphisms (SNPs) to exactly one of the two chromosomes. Indeed, the knowledge of complete haplotypes is generally more informative than analyzing single SNPs and plays a fundamental role in many medical applications. RESULTS: To reconstruct the two haplotypes, we addressed the weighted Minimum Error Correction (wMEC) problem, which is a successful approach for haplotype assembly. This NP-hard problem consists in computing the two haplotypes that partition the sequencing reads into two disjoint sub-sets, with the least number of corrections to the SNP values. To this aim, we propose here GenHap, a novel computational method for haplotype assembly based on Genetic Algorithms, yielding optimal solutions by means of a global search process. In order to evaluate the effectiveness of our approach, we run GenHap on two synthetic (yet realistic) datasets, based on the Roche/454 and PacBio RS II sequencing technologies. We compared the performance of GenHap against HapCol, an efficient state-of-the-art algorithm for haplotype phasing. Our results show that GenHap always obtains high accuracy solutions (in terms of haplotype error rate), and is up to 4× faster than HapCol in the case of Roche/454 instances and up to 20× faster when compared on the PacBio RS II dataset. Finally, we assessed the performance of GenHap on two different real datasets. CONCLUSIONS: Future-generation sequencing technologies, producing longer reads with higher coverage, can highly benefit from GenHap, thanks to its capability of efficiently solving large instances of the haplotype assembly problem. Moreover, the optimization approach proposed in GenHap can be extended to the study of allele-specific genomic features, such as expression, methylation and chromatin conformation, by exploiting multi-objective optimization techniques. The source code and the full documentation are available at the following GitHub repository: https://github.com/andrea-tango/GenHap .

Gut microbiota and human relationships are strictly connected to each other. What we eat reflects our body-mind connection and synchronizes with people around us. However, how this impacts on gut microbiota and, conversely, how gut bacteria influence our dietary behaviors has not been explored yet. To quantify the complex dynamics of this interplay between gut and human behaviors we explore the "gut-human behavior axis" and its evolutionary dynamics in a real-world scenario represented by the social multiplex network. We consider a dual type of similarity, homophily and gut similarity, other than psychological and unconscious biases. We analyze the dynamics of social and gut microbial communities, quantifying the impact of human behaviors on diets and gut microbial composition and, backwards, through a control mechanism. Meal timing mechanisms and "chrono-nutrition" play a crucial role in feeding behaviors, along with the quality and quantity of food intake. Considering a population of shift workers, we explore the dynamic interplay between their eating behaviors and gut microbiota, modeling the social dynamics of chrono-nutrition in a multiplex network. Our findings allow us to quantify the relation between human behaviors and gut microbiota through the methodological introduction of gut metabolic modeling and statistical estimators, able to capture their dynamic interplay. Moreover, we find that the timing of gut microbial communities is slower than social interactions and shift-working, and the impact of shift-working on the dynamics of chrono-nutrition is a fluctuation of strategies with a major propensity for defection (e.g. high-fat meals). A deeper understanding of the relation between gut microbiota and the dietary behavioral patterns, by embedding also the related social aspects, allows improving the overall knowledge about metabolic models and their implications for human health, opening the possibility to design promising social therapeutic dietary interventions.

Along with the rapid development of the Internet, people more incline to access to the network for life needs through intelligent terminal. Once the devices are beyond control, the privacy information are leaked. Hence it's important to protect the data obtained by client. From above view, a terminal sensitive data protection method by adjusting access time automatically is proposed. This method achieves fine grit visit control based on sliding time window, applies dynamic authorization rules based on temporary parameters, employs temporary parameters to identify users and control their visits, and uses the data desensitization to protect sensitive privacy. It can ensure the security of obtained data by client, and also meets the availability of the data. The proposed method employing access control, time authorization and desensitization was compared with those existing data protection methods, as well as its validity was verified.

A small variation in mobile hardware and software can potentially cause a significant heterogeneity or variation in the sensor data each device collects. For example, the microphone and accelerometer sensors on different devices can respond very differently to the same audio or motion phenomena. Other factors, like the instantaneous computational load on a smartphone, can cause key behavior like sensor sampling rates to fluctuate, further polluting the data. When sensing devices are deployed in unconstrained and real-world conditions, examples of sharply lower classification accuracy are observed due to what is collectively known as the sensing system heterogeneity. In this work, we take an unconventional approach and argue against solving individual forms of heterogeneity, e.g., improving OS behavior, or the quality/uniformity of components. Instead, we propose and build classifiers that themselves are more tolerant of these variations by leveraging deep learning and a data-augmented training process. Neither augmentation nor deep learning has previously been attempted to cope with sensor heterogeneity. We systematically investigate how these two machine learning methodologies can be adapted to solve such problems, and identify when and where they are able to be successful. We find that our proposed approach is able to reduce classifier errors on an average by 9% and 17% for a range of inertial-and audio-based mobile classification tasks.

Metabolic modelling has entered a mature phase with dozens of methods and software implementations available to the practitioner and the theoretician. It is not easy for a modeller to be able to see the wood (or the forest) for the trees. Driven by this analogy, we here present a 'forest' of principal methods used for constraint-based modelling in systems biology. This provides a tree-based view of methods available to prospective modellers, also available in interactive version at http://modellingmetabolism.net, where it will be kept updated with new methods after the publication of the present manuscript. Our updated classification of existing methods and tools highlights the most promising in the different branches, with the aim to develop a vision of how existing methods could hybridize and become more complex. We then provide the first hands-on tutorial for multi-objective optimization of metabolic models in R. We finally discuss the implementation of multi-view machine learning approaches in poly-omic integration. Throughout this work, we demonstrate the optimization of trade-offs between multiple metabolic objectives, with a focus on omic data integration through machine learning. We anticipate that the combination of a survey, a perspective on multi-view machine learning and a step-by-step R tutorial should be of interest for both the beginner and the advanced user.

Motivation: Antibodies play essential roles in the immune system of vertebrates and are powerful tools in research and diagnostics. While hypervariable regions of antibodies, which are responsible for binding, can be readily identified from their amino acid sequence, it remains challenging to accurately pinpoint which amino acids will be in contact with the antigen (the paratope). Results: In this work, we present a sequence-based probabilistic machine learning algorithm for paratope prediction, named Parapred. Parapred uses a deep-learning architecture to leverage features from both local residue neighbourhoods and across the entire sequence. The method significantly improves on the current state-of-the-art methodology, and only requires a stretch of amino acid sequence corresponding to a hypervariable region as an input, without any information about the antigen. We further show that our predictions can be used to improve both speed and accuracy of a rigid docking algorithm. Availability and implementation: The Parapred method is freely available as a webserver at http://www-mvsoftware.ch.cam.ac.uk/and for download at https://github.com/eliberis/parapred. Supplementary information: Supplementary information is available at Bioinformatics online.

BACKGROUND: Two important challenges in the analysis of molecular biology information are data (multi-omic information) integration and the detection of patterns across large scale molecular networks and sequences. They are are actually coupled beause the integration of omic information may provide better means to detect multi-omic patterns that could reveal multi-scale or emerging properties at the phenotype levels. RESULTS: Here we address the problem of integrating various types of molecular information (a large collection of gene expression and sequence data, codon usage and protein abundances) to analyse the E.coli metabolic response to treatments at the whole network level. Our algorithm, MORA (Multi-omic relations adjacency) is able to detect patterns which may represent metabolic network motifs at pathway and supra pathway levels which could hint at some functional role. We provide a description and insights on the algorithm by testing it on a large database of responses to antibiotics. Along with the algorithm MORA, a novel model for the analysis of oscillating multi-omics has been proposed. Interestingly, the resulting analysis suggests that some motifs reveal recurring oscillating or position variation patterns on multi-omics metabolic networks. Our framework, implemented in R, provides effective and friendly means to design intervention scenarios on real data. By analysing how multi-omics data build up multi-scale phenotypes, the software allows to compare and test metabolic models, design new pathways or redesign existing metabolic pathways and validate in silico metabolic models using nearby species. CONCLUSIONS: The integration of multi-omic data reveals that E.coli multi-omic metabolic networks contain position dependent and recurring patterns which could provide clues of long range correlations in the bacterial genome.

BACKGROUND: De novo assembly of RNA-seq data allows the study of transcriptome in absence of a reference genome either if data is obtained from a single organism or from a mixed sample as in metatranscriptomics studies. Given the high number of sequences obtained from NGS approaches, a critical step in any analysis workflow is the assembly of reads to reconstruct transcripts thus reducing the complexity of the analysis. Despite many available tools show a good sensitivity, there is a high percentage of false positives due to the high number of assemblies considered and it is likely that the high frequency of false positive is underestimated by currently used benchmarks. The reconstruction of not existing transcripts may false the biological interpretation of results as - for example - may overestimate the identification of "novel" transcripts. Moreover, benchmarks performed are usually based on RNA-seq data from annotated genomes and assembled transcripts are compared to annotations and genomes to identify putative good and wrong reconstructions, but these tests alone may lead to accept a particular type of false positive as true, as better described below. RESULTS: Here we present a novel methodology of de novo assembly, implemented in a software named STAble (Short-reads Transcriptome Assembler). The novel concept of this assembler is that the whole reads are used to determine possible alignments instead of using smaller k-mers, with the aim of reducing the number of chimeras produced. Furthermore, we applied a new set of benchmarks based on simulated data to better define the performance of assembly method and carefully identifying true reconstructions. STAble was also used to build a prototype workflow to analyse metatranscriptomics data in connection to a steady state metabolic modelling algorithm. This algorithm was used to produce high quality metabolic interpretations of small gene expression sets obtained from already published RNA-seq data that we assembled with STAble. CONCLUSIONS: The presented results, albeit preliminary, clearly suggest that with this approach is possible to identify informative reactions not directly revealed by raw transcriptomic data.

BACKGROUND: Inflammation is a core element of many different, systemic and chronic diseases that usually involve an important autoimmune component. The clinical phase of inflammatory diseases is often the culmination of a long series of pathologic events that started years before. The systemic characteristics and related mechanisms could be investigated through the multi-omic comparative analysis of many inflammatory diseases. Therefore, it is important to use molecular data to study the genesis of the diseases. Here we propose a new methodology to study the relationships between inflammatory diseases and signalling molecules whose dysregulation at molecular levels could lead to systemic pathological events observed in inflammatory diseases. RESULTS: We first perform an exploratory analysis of gene expression data of a number of diseases that involve a strong inflammatory component. The comparison of gene expression between disease and healthy samples reveals the importance of members of gene families coding for signalling factors. Next, we focus on interested signalling gene families and a subset of inflammation related diseases with multi-omic features including both gene expression and DNA methylation. We introduce a phylogenetic-based multi-omic method to study the relationships between multi-omic features of inflammation related diseases by integrating gene expression, DNA methylation through sequence based phylogeny of the signalling gene families. The models of adaptations between gene expression and DNA methylation can be inferred from pre-estimated evolutionary relationship of a gene family. Members of the gene family whose expression or methylation levels significantly deviate from the model are considered as the potential disease associated genes. CONCLUSIONS: Applying the methodology to four gene families (the chemokine receptor family, the TNF receptor family, the TGF- β gene family, the IL-17 gene family) in nine inflammation related diseases, we identify disease associated genes which exhibit significant dysregulation in gene expression or DNA methylation in the inflammation related diseases, which provides clues for functional associations between the diseases.

The eight papers in this special section were presented at the 14th International Conference on Computational Methods in Systems Biology (CMSB 2016)that was held at the Computer Laboratory, University of Cambridge, UK, on September 21-23, 2016.

The outbreak of the Ebola virus in recent years has resulted in numerous research initiatives to seek new solutions to contain the virus. A number of approaches that have been investigated include new vaccines to boost the immune system. An alternative post-exposure treatment is presented in this paper. The proposed approach for clearing the Ebola virus can be developed through a microfluidic attenuator, which contains the engineered bacteria that traps Ebola flowing through the blood onto its membrane. The paper presents the analysis of the chemical binding force between the virus and a genetically engineered bacterium considering the opposing forces acting on the attachment point, including hydrodynamic tension and drag force. To test the efficacy of the technique, simulations of bacterial motility within a confined area to trap the virus were performed. More than 60 percent of the displaced virus could be collected within 15 minutes. While the proposed approach currently focuses on in vitro environments for trapping the virus, the system can be further developed into a future treatment system whereby blood can be cycled out of the body into a microfluidic device that contains the engineered bacteria to trap viruses.

We introduce a novel Bayesian hybrid matrix factorisation model (HMF) for data integration, based on combining multiple matrix factorisation methods, that can be used for in- and out-of-matrix prediction of missing values. The model is very general and can be used to integrate many datasets across different entity types, including repeated experiments, similarity matrices, and very sparse datasets. We apply our method on two biological applications, and extensively compare it to state-of-the-art machine learning and matrix factorisation models. For in-matrix predictions on drug sensitivity datasets we obtain consistently better performances than existing methods. This is especially the case when we increase the sparsity of the datasets. Furthermore, we perform out-of-matrix predictions on methylation and gene expression datasets, and obtain the best results on two of the three datasets, especially when the predictivity of datasets is high.

Precision and personalized medicine will be increasingly based on the integration of various type of information, particularly electronic health records and genome sequences. The availability of cheap genome sequencing services and the information interoperability will increase the role of online bioinformatics analysis. Being on the Internet poses constant threats to security and privacy. While we are connected and we share information, websites and internet services collect various types of personal data with or without the user consent. It is likely that genomics will merge with the internet culture of connectivity. This process will increase incidental findings, exposure and vulnerability. Here we discuss the social vulnerability owing to the genome and Internet combined security and privacy weaknesses. This urges more efforts in education and social awareness on how biomedical data are analysed and transferred through the internet and how inferential methods could integrate information from different sources. We propose that digital social platforms, used for raising collective awareness in different fields, could be developed for collaborative and bottom-up efforts in education. In this context, bioinformaticians could play a meaningful role in mitigating the future risk of digital-genomic divide.

Nervous cells and blood vessels form crucial circulating networks in human body. They are interdependent on biological level and communicate with each other via abundant interaction phenomena. These interactions are complex while worthy concerned, which may guide to implement controllable relaying communication across nervous and blood vascular heterogeneous channels. In this paper, we highlight the heterogeneous theme in area of molecular communication. We set up a basic framework based on the heterogeneous network interactions, in which two properties are proposed to make those interactions easier understood on communication level. Moreover, we establish a one-way single-threaded channel model as a case study based on the network framework, and give the mutual information expressions. We aim to explore the possibility of effective communication. The results show that settings of the adjustable system have a significant impact on the relaying performance, as well as the mutual information.

Systems Medicine (SM) can be defined as an extension of Systems Biology (SB) to Clinical-Epidemiological disciplines through a shifting paradigm, starting from a cellular, toward a patient centered framework. According to this vision, the three pillars of SM are Biomedical hypotheses, experimental data, mainly achieved by Omics technologies and tailored computational, statistical and modeling tools. The three SM pillars are highly interconnected, and their balancing is crucial. Despite the great technological progresses producing huge amount of data (Big Data) and impressive computational facilities, the Bio-Medical hypotheses are still of primary importance. A paradigmatic example of unifying Bio-Medical theory is the concept of Inflammaging. This complex phenotype is involved in a large number of pathologies and patho-physiological processes such as aging, age-related diseases and cancer, all sharing a common inflammatory pathogenesis. This Biomedical hypothesis can be mapped into an ecological perspective capable to describe by quantitative and predictive models some experimentally observed features, such as microenvironment, niche partitioning and phenotype propagation. In this article we show how this idea can be supported by computational methods useful to successfully integrate, analyze and model large data sets, combining cross-sectional and longitudinal information on clinical, environmental and omics data of healthy subjects and patients to provide new multidimensional biomarkers capable of distinguishing between different pathological conditions, e.g. healthy versus unhealthy state, physiological versus pathological aging.

Chromosome conformation capture techniques are producing a huge amount of data about the architecture of our genome. These data can provide us with a better understanding of the events that induce critical regulations of the cellular function from small changes in the three-dimensional genome architecture. Generating a unified view of spatial, temporal, genetic and epigenetic properties poses various challenges of data analysis, visualization, integration and mining, as well as of high performance computing and big data management. Here, we describe the critical issues of this new branch of bioinformatics, oriented at the comprehension of the three-dimensional genome architecture, which we call 'Nucleome Bioinformatics', looking beyond the currently available tools and methods, and highlight yet unaddressed challenges and the potential approaches that could be applied for tackling them. Our review provides a map for researchers interested in using computer science for studying 'Nucleome Bioinformatics', to achieve a better understanding of the biological processes that occur inside the nucleus.

MOTIVATION: Recent advancements in molecular methods have made it possible to capture physical contacts between multiple chromatin fragments. The resulting association matrices provide a noisy estimate for average spatial proximity that can be used to gain insights into the genome organization inside the nucleus. However, extracting topological information from these data is challenging and their integration across resolutions is still poorly addressed. Recent findings suggest that a hierarchical approach could be advantageous for addressing these challenges. RESULTS: We present an algorithmic framework, which is based on hierarchical block matrices (HBMs), for topological analysis and integration of chromosome conformation capture (3C) data. We first describe chromoHBM, an algorithm that compresses high-throughput 3C (HiT-3C) data into topological features that are efficiently summarized with an HBM representation. We suggest that instead of directly combining HiT-3C datasets across resolutions, which is a difficult task, we can integrate their HBM representations, and describe chromoHBM-3C, an algorithm which merges HBMs. Since three-dimensional (3D) reconstruction can also benefit from topological information, we further present chromoHBM-3D, an algorithm which exploits the HBM representation in order to gradually introduce topological constraints to the reconstruction process. We evaluate our approach in light of previous image microscopy findings and epigenetic data, and show that it can relate multiple spatial scales and provide a more complete view of the 3D genome architecture. AVAILABILITY AND IMPLEMENTATION: The presented algorithms are available from: https://github.com/yolish/hbm CONTACT: [email protected] or [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: With the development of experimental methods and technology, we are able to reliably gain access to data in larger quantities, dimensions and types. This has great potential for the improvement of machine learning (as the learning algorithms have access to a larger space of information). However, conventional machine learning approaches used thus far on single-dimensional data inputs are unlikely to be expressive enough to accurately model the problem in higher dimensions; in fact, it should generally be most suitable to represent our underlying models as some form of complex networksng;nsio with nontrivial topological features. As the first step in establishing such a trend, we present MUXSTEP: , an open-source library utilising multiplex networks for the purposes of binary classification on multiple data types. The library is designed to be used out-of-the-box for developing models based on the multiplex network framework, as well as easily modifiable to suit problem modelling needs that may differ significantly from the default approach described. AVAILABILITY AND IMPLEMENTATION: The full source code is available on GitHub: https://github.com/PetarV-/muxstep CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Genomic, transcriptomic, and metabolic variations shape the complex adaptation landscape of bacteria to varying environmental conditions. Elucidating the genotype-phenotype relation paves the way for the prediction of such effects, but methods for characterizing the relationship between multiple environmental factors are still lacking. Here, we tackle the problem of extracting network-level information from collections of environmental conditions, by integrating the multiple omic levels at which the bacterial response is measured. RESULTS: To this end, we model a large compendium of growth conditions as a multiplex network consisting of transcriptomic and fluxomic layers, and we propose a multi-omic network approach to infer similarity of growth conditions by integrating layers of the multiplex network. Each node of the network represents a single condition, while edges are similarities between conditions, as measured by phenotypic and transcriptomic properties on different layers of the network. We then fuse these layers into one network, therefore capturing a global network of conditions and the associated similarities across two omic levels. We apply this multi-omic fusion to an updated genome-scale reconstruction of Escherichia coli that includes underground metabolism and new gene-protein-reaction associations. CONCLUSIONS: Our method can be readily used to evaluate and cross-compare different collections of conditions among different species. Acquiring multi-omic information on the topology of the space of experimental conditions makes it possible to infer the position and to build condition-specific models of untested or incomplete profiles for which experimental data is not available. Our weighted network fusion method for genome-scale models is freely available at https://github.com/maxconway/SNFtool .

As the amount of biological data in the public domain grows, so does the range of modeling and analysis techniques employed in systems biology. In recent years, a number of theoretical computer science developments have enabled modeling methodology to keep pace. The growing interest in systems biology in executable models and their analysis has necessitated the borrowing of terms and methods from computer science, such as formal analysis, model checking, static analysis, and runtime verification. Here, we discuss the most important and exciting computational methods and tools currently available to systems biologists. We believe that a deeper understanding of the concepts and theory highlighted in this review will produce better software practice, improved investigation of complex biological processes, and even new ideas and better feedback into computer science.

We are often able to describe what happens at almost all biological length scales, from the molecular level to the whole organism; however, putting things together in order to obtain real comprehension is difficult and less developed. The challenge is to develop novel computational intelligence frameworks that integrate the different layers of molecular information. The introduction of such methodologies would enable to discover the relation between the environmental (external) conditions and the changes in the metabolic multi omic networks (i.e. the adaptive response of the internal environment). Many molecular levels can contribute to adaptability: pathways structure, codon usage bias, transcriptomics and metabolism. Here, we develop a method that combines Probabilistic Suffix Trees and Clustering models to integrate bacterial molecular information and to map different conditions to an omic multi-dimensional objective space. This methodology allows to identify oscillations in network structure of metabolic pathways. We tested our method by considering two case studies (from Escherichia coli) with antibiotics added to the medium. These oscillations provide insights into novel emerging properties of multi omic networks and to a better understanding of antibiotic effects and their gene targeting constraints.

High-throughput molecular biology techniques are widely used to identify physical interactions between genetic elements located throughout the human genome. Chromosome Conformation Capture (3C) and other related techniques allow to investigate the spatial organisation of chromosomes in the cell's natural state. Recent results have shown that there is a large correlation between co-localization and co-regulation of genes, but these important information are hampered by the lack of biologists-friendly analysis and visualisation software. In this work we introduce NuChart-II, a tool for Hi-C data analysis that provides a gene-centric view of the chromosomal neighbourhood in a graph-based manner. NuChart-II is an efficient and highly optimized C++ re-implementation of a previous prototype package developed in R. Representing Hi-C data using a graph-based approach overcomes the common view relying on genomic coordinates and permits the use of graph analysis techniques to explore the spatial conformation of a gene neighbourhood.

Artificial Life (hereafter ALife) is an interdisciplinary research field that has the purpose of gaining a better understanding of biological life by artificially synthesizing simple and novel forms of life, as well as reproducing lifelike properties of living systems. Synthesis of artificial cells, simulation of large-scale biological networks, intelligent use of exponentially growing amounts of biochemical data, exploitation of biological substrates for computation and control, and deployment of bio-inspired engineering are just some of the main issues of ALife and nowadays are cutting-edge topics.ALife is at the intersection between a theoretical perspective, namely, the scientific explanations of different levels of life organization (e.g., molecules, compartments, cells, tissues, organs, organisms, societies, and collective and social phenomena) and advanced technological applications (bio-inspired algorithms and techniques for building up effective solutions such as in the fields of robotics, big data analysis, and computational medicine).This special issue presents the top six articles—revised and extended—that were presented at ECAL 2013, the twelfth European Conference on Artificial Life, which was held in Taormina, Sicily, Italy, September 2–6, 2013 (http://www.dmi.unict.it/ecal2013/). ECAL 2013 was a grand scientific celebration with hundreds of paper and poster presentations (for a total of 267 submissions, 10 workshops, 16 tracks, and more than 350 attendants) by leading scientists in the field, which have described an impressive array of results, ideas, technologies, and applications, thus showing the current state of the art of ALife. The aim of this special issue is to give an incentive for boosting interest and new ideas in designing life and lifelike processes at different levels of complexity.After a careful peer-review process, only the top six research manuscripts, extended and revised, of ECAL 2013 have been selected for inclusion in this special issue.In “Multi-crease Self-folding by Global Heating,” by Shuhei Miyashita, Cagdas Onal, and Daniela Rus, the authors present a scheme for the autonomous folding of the body of a 3D robot from a 2D sheet using heat. They developed an easy, fast, and reliable fabrication technique for constructing a self-folding sheet. In particular, they exploit the thermal deformation of a constructive sheet sandwich made of rigid structural layers, and they utilize the V-fold method to attain the targeted folding angles. The authors also developed a mobile robot that uses such a self-folding body and that performs locomotion by using two vibration motors. The locomotion of the robot proves the functionality of the self-folding origami structure.In “The Search for Candidate Relevant Subsets of Variables in Complex Systems,” by M. Villani, A. Roli, A. Filisetti, M. Fiorucci, I. Poli, and R. Serra, the authors present a method to identify relevant subsets of variables for the comprehension of the organization of a dynamical system. Thus, the dynamical cluster index (DCI)—an information-theoretic measure—is introduced; the DCI relies on observations of current values of the relationships among the system variables, and does not require any previous knowledge. The usefulness of this method was tested in different application domains, where a known dynamical model generates the data, and the aim of the DCI is to uncover significant aspects of the organization of the model. One of the main novelties of this work is in the use of truly dynamical systems.In “Indirectly Encoding Running and Jumping Sodarace Creatures for Artificial Life,” by Paul Szerlip and Kenneth O. Stanley, the authors present a platform for evolving two-dimensional artificial creatures, whose aim is to serve for future artificial life experiments in evolving creatures. They introduce a new, indirectly encoded Sodarace (IESoR) system, which extends the original Sodarace by enabling the evolution of significantly more complex and regular creature morphologies. The capability of the system has been proved in both walking and jumping domains, in which IESoR discovered a wide breadth of strategies through the novel approach of search with local competition.In “Experiments on and Numerical Modeling of the Capture and Concentration of Transcription-Translation Machinery inside Vesicles,” by Fabio Mavelli and Pasquale Stano, the authors present a mathematical model of the encapsulation of transcription-translation (TX-TL) coupled reactions based on a minimal protein synthesis model and on different solute partition functions. Such a model can be used to predict the time span and the amount of protein produced starting from any pure system composition. The proposed approach highlights the role of stochastic events in synthetic-cell research, and emphasizes the importance of the integration of stochastic simulations with experimental approaches. The results presented show clearly that experimental data are compatible with an entrapment model, which follows a power law rather than a Gaussian distribution.In “Lessons from Speciation Dynamics: How to Generate Selective Pressure Towards Diversity,” by Heiko Hamann, the author detects how methods for generating selective pressure towards diversity (SPTD) can be transferred from the domain of artificial ecology (AE) to evolutionary robotics (ER). Furthermore, he also investigates how SPTD is generated without task-specific behavioral features or other forms of a priori knowledge. A promising finding of this work is that selective pressure towards unpopulated regions of the search space is generated for both systems. In particular, a beneficial finding is the analogy between the behavioral distances in ER and self-organizing ecology, in terms of the selective pressure generated.In “Cell-Division Behavior in a Heterogeneous Swarm Environment,” by Adam Erskine and J. Michael Herrmann, the authors present a system of virtual particles that interact using simple kinetic rules. In particular, they present a two-species three-dimensional swarm in which the behavior that emerges resembles cell division. In addition, they have proved that behavior resembling repeated cell division emerges from the low-level kinetic interactions of a heterogeneous swarm. Such division behavior exists in a small but finite volume of the swarm chemistry parameter space. Furthermore, from the outcomes it emerges how the behaviors differ depending on whether the swarm is moving in two- or three-dimensional space.The guest editors would like to thank all the authors of this special issue for their excellent work, constituting a valuable contribution to the state of the art and the scientific growth of ALife. We would also like to thank everyone who helped to ensure high scientific quality both of ECAL 2013 and of this special issue: the plenary speakers, Prof. Paolo Arena, Prof. Roberto Cingolani, Prof. Roberto Cipolla, Prof. Martin Hanczyc, Prof. Henrik Hautop Lund, Prof. Didier Keymeulen, Prof. Steve Oliver, Prof. Bernhard Ø. Palsson, and Prof. Rolf Pfeifer; all technical program committee members; the workshop chair; the tutorial chair; the publicity chair; and the local organizers. Special thanks go to the editor-in-chief of Artificial Life, Prof. Mark Bedau, for encouraging and accepting this special issue, and to the editorial assistant, Linda Reedijk, for her useful and important support in the realization of this volume.

Abstract Motivation: Completing the genome sequence of an organism is an important task in comparative, functional and structural genomics. However, this remains a challenging issue from both a computational and an experimental viewpoint. Genome scaffolding (i.e. the process of ordering and orientating contigs) of de novo assemblies usually represents the first step in most genome finishing pipelines. Results: In this article we present MeDuSa (Multi-Draft based Scaffolder), an algorithm for genome scaffolding. MeDuSa exploits information obtained from a set of (draft or closed) genomes from related organisms to determine the correct order and orientation of the contigs. MeDuSa formalizes the scaffolding problem by means of a combinatorial optimization formulation on graphs and implements an efficient constant factor approximation algorithm to solve it. In contrast to currently used scaffolders, it does not require either prior knowledge on the microrganisms dataset under analysis (e.g. their phylogenetic relationships) or the availability of paired end read libraries. This makes usability and running time two additional important features of our method. Moreover, benchmarks and tests on real bacterial datasets showed that MeDuSa is highly accurate and, in most cases, outperforms traditional scaffolders. The possibility to use MeDuSa on eukaryotic datasets has also been evaluated, leading to interesting results. Availability and implementation: MeDuSa web server: http://combo.dbe.unifi.it/medusa. A stand-alone version of the software can be downloaded from https://github.com/combogenomics/medusa/releases. All results presented in this work have been obtained with MeDuSa v. 1.3. Contact: [email protected] Supplementary information: Supplementary data are available at Bioinformatics online.

CytoCom is an interactive plugin for Cytoscape that can be used to search, explore, analyse and visualize human disease comorbidity network. It represents disease-disease associations in terms of bipartite graphs and provides International Classification of Diseases, Ninth Revision (ICD9)-centric and disease name centric views of disease information. It allows users to find associations between diseases based on the two measures: Relative Risk (RR) and [Formula: see text]-correlation values. In the disease network, the size of each node is based on the prevalence of that disease. CytoCom is capable of clustering disease network based on the ICD9 disease category. It provides user-friendly access that facilitates exploration of human diseases, and finds additional associated diseases by double-clicking a node in the existing network. Additional comorbid diseases are then connected to the existing network. It is able to assist users for interpretation and exploration of the human diseases by a variety of built-in functions. Moreover, CytoCom permits multi-colouring of disease nodes according to standard disease classification for expedient visualization.

Ductal carcinoma is one of the most common cancers among women, and the main cause of death is the formation of metastases. The development of metastases is caused by cancer cells that migrate from the primary tumour site (the mammary duct) through the blood vessels and extravasating they initiate metastasis. Here, we propose a multi-compartment model which mimics the dynamics of tumoural cells in the mammary duct, in the circulatory system and in the bone. Through a branching process model, we describe the relation between the survival times and the four markers mainly involved in metastatic breast cancer (EPCAM, CD47, CD44 and MET). In particular, the model takes into account the gene expression profile of circulating tumour cells to predict personalised survival probability. We also include the administration of drugs as bisphosphonates, which reduce the formation of circulating tumour cells and their survival in the blood vessels, in order to analyse the dynamic changes induced by the therapy. We analyse the effects of circulating tumour cells on the progression of the disease providing a quantitative measure of the cell driver mutations needed for invading the bone tissue. Our model allows to design intervention scenarios that alter the patient-specific survival probability by modifying the populations of circulating tumour cells and it could be extended to other cancer metastasis dynamics.

UNLABELLED: The fluorescence in situ hybridization (FISH) method has been providing valuable information on physical distances between loci (via image analysis) for several decades. Recently, high-throughput data on nearby chemical contacts between and within chromosomes became available with the Hi-C method. Here, we present FisHiCal, an R package for an iterative FISH-based Hi-C calibration that exploits in full the information coming from these methods. We describe here our calibration model and present 3D inference methods that we have developed for increasing its usability, namely, 3D reconstruction through local stress minimization and detection of spatial inconsistencies. We next confirm our calibration across three human cell lines and explain how the output of our methods could inform our model, defining an iterative calibration pipeline, with applications for quality assessment and meta-analysis. AVAILABILITY AND IMPLEMENTATION: FisHiCal v1.1 is available from http://cran.r-project.org/.

BACKGROUND: Infections are often associated to comorbidity that increases the risk of medical conditions which can lead to further morbidity and mortality. SARS is a threat which is similar to MERS virus, but the comorbidity is the key aspect to underline their different impacts. One UK doctor says "I'd rather have HIV than diabetes" as life expectancy among diabetes patients is lower than that of HIV. However, HIV has a comorbidity impact on the diabetes. RESULTS: We present a quantitative framework to compare and explore comorbidity between diseases. By using neighbourhood based benchmark and topological methods, we have built comorbidity relationships network based on the OMIM and our identified significant genes. Then based on the gene expression, PPI and signalling pathways data, we investigate the comorbidity association of these 2 infective pathologies with other 7 diseases (heart failure, kidney disorder, breast cancer, neurodegenerative disorders, bone diseases, Type 1 and Type 2 diabetes). Phenotypic association is measured by calculating both the Relative Risk as the quantified measures of comorbidity tendency of two disease pairs and the ϕ-correlation to measure the robustness of the comorbidity associations. The differential gene expression profiling strongly suggests that the response of SARS affected patients seems to be mainly an innate inflammatory response and statistically dysregulates a large number of genes, pathways and PPIs subnetworks in different pathologies such as chronic heart failure (21 genes), breast cancer (16 genes) and bone diseases (11 genes). HIV-1 induces comorbidities relationship with many other diseases, particularly strong correlation with the neurological, cancer, metabolic and immunological diseases. Similar comorbidities risk is observed from the clinical information. Moreover, SARS and HIV infections dysregulate 4 genes (ANXA3, GNS, HIST1H1C, RASA3) and 3 genes (HBA1, TFRC, GHITM) respectively that affect the ageing process. It is notable that HIV and SARS similarly dysregulated 11 genes and 3 pathways. Only 4 significantly dysregulated genes are common between SARS-CoV and MERS-CoV, including NFKBIA that is a key regulator of immune responsiveness implicated in susceptibility to infectious and inflammatory diseases. CONCLUSIONS: Our method presents a ripe opportunity to use data-driven approaches for advancing our current knowledge on disease mechanism and predicting disease comorbidities in a quantitative way.

We propose a framework to design metabolic pathways in which many objectives are optimized simultaneously. This allows to characterize the energy signature in models of algal and mitochondrial metabolism. The optimal design and assessment of the model is achieved through a multi-objective optimization technique driven by epsilon-dominance and identifiability analysis. A faster convergence process with robust candidate solutions is permitted by a relaxed Pareto dominance, regulating the granularity of the approximation of the Pareto front. Our framework is also suitable for black-box analysis, enabling to investigate and optimize any biological pathway modeled with ODEs, DAEs, FBA and GPR.

SUMMARY: With the introduction of the Hi-C method new and fundamental properties of the nuclear architecture are emerging. The ability to interpret data generated by this method, which aims to capture the physical proximity between and within chromosomes, is crucial for uncovering the three dimensional structure of the nucleus. Providing researchers with tools for interactive visualization of Hi-C data can help in gaining new and important insights. Specifically, visual comparison can pinpoint changes in spatial organization between Hi-C datasets, originating from different cell lines or different species, or normalized by different methods. Here, we present CytoHiC, a Cytsocape plugin, which allow users to view and compare spatial maps of genomic landmarks, based on normalized Hi-C datasets. CytoHiC was developed to support intuitive visual comparison of Hi-C data and integration of additional genomic annotations. AVAILABILITY: The CytoHiC plugin, source code, user manual, example files and documentation are available at: http://apps.cytoscape.org/apps/cytohicplugin

In low and high eukaryotes, energy is collected or transformed in compartments, the organelles. The rich variety of size, characteristics, and density of the organelles makes it difficult to build a general picture. In this paper, we make use of the Pareto-front analysis to investigate the optimization of energy metabolism in mitochondria and chloroplasts. Using the Pareto optimality principle, we compare models of organelle metabolism on the basis of single- and multiobjective optimization, approximation techniques (the Bayesian Automatic Relevance Determination), robustness, and pathway sensitivity analysis. Finally, we report the first analysis of the metabolic model for the hydrogenosome of Trichomonas vaginalis, which is found in several protozoan parasites. Our analysis has shown the importance of the Pareto optimality for such comparison and for insights into the evolution of the metabolism from cytoplasmic to organelle bound, involving a model order reduction. We report that Pareto fronts represent an asymptotic analysis useful to describe the metabolism of an organism aimed at maximizing concurrently two or more metabolite concentrations.

MOTIVATION: Metabolic engineering algorithms provide means to optimize a biological process leading to the improvement of a biotechnological interesting molecule. Therefore, it is important to understand how to act in a metabolic pathway in order to have the best results in terms of productions. In this work, we present a computational framework that searches for optimal and robust microbial strains that are able to produce target molecules. Our framework performs three tasks: it evaluates the parameter sensitivity of the microbial model, searches for the optimal genetic or fluxes design and finally calculates the robustness of the microbial strains. We are capable to combine the exploration of species, reactions, pathways and knockout parameter spaces with the Pareto-optimality principle. RESULTS: Our framework provides also theoretical and practical guidelines for design automation. The statistical cross comparison of our new optimization procedures, performed with respect to currently widely used algorithms for bacteria (e.g. Escherichia coli) over different multiple functions, reveals good performances over a variety of biotechnological products. AVAILABILITY: http://www.dmi.unict.it/nicosia/pathDesign.html. CONTACT: [email protected] or [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: This work focuses on the computational modelling of osteomyelitis, a bone pathology caused by bacteria infection (mostly Staphylococcus aureus). The infection alters the RANK/RANKL/OPG signalling dynamics that regulates osteoblasts and osteoclasts behaviour in bone remodelling, i.e. the resorption and mineralization activity. The infection rapidly leads to severe bone loss, necrosis of the affected portion, and it may even spread to other parts of the body. On the other hand, osteoporosis is not a bacterial infection but similarly is a defective bone pathology arising due to imbalances in the RANK/RANKL/OPG molecular pathway, and due to the progressive weakening of bone structure. RESULTS: Since both osteoporosis and osteomyelitis cause loss of bone mass, we focused on comparing the dynamics of these diseases by means of computational models. Firstly, we performed meta-analysis on a gene expression data of normal, osteoporotic and osteomyelitis bone conditions. We mainly focused on RANKL/OPG signalling, the TNF and TNF receptor superfamilies and the NF-kB pathway. Using information from the gene expression data we estimated parameters for a novel model of osteoporosis and of osteomyelitis. Our models could be seen as a hybrid ODE and probabilistic verification modelling framework which aims at investigating the dynamics of the effects of the infection in bone remodelling. Finally we discuss different diagnostic estimators defined by formal verification techniques, in order to assess different bone pathologies (osteopenia, osteoporosis and osteomyelitis) in an effective way. CONCLUSIONS: We present a modeling framework able to reproduce aspects of the different bone remodeling defective dynamics of osteomyelitis and osteoporosis. We report that the verification-based estimators are meaningful in the light of a feed forward between computational medicine and clinical bioinformatics.

Our work focuses on bone remodeling with a multiscale breadth that ranges from modeling intracellular and intercellular RANK/RANKL signaling to tissue dynamics, by developing a multilevel modeling framework. Several important findings provide clear evidences of the multiscale properties of bone formation and of the links between RANK/RANKL and bone density in healthy and disease conditions. Recent studies indicate that the circulating levels of OPG and RANKL are inversely related to bone turnover and Bone Mineral Density (BMD) and contribute to the development of osteoporosis in postmenopausal women, and thalassemic patients. We make use of a spatial process algebra, the Shape Calculus, to control stochastic cell agents that are continuously remodeling the bone. We found that our description is effective for such a multiscale, multilevel process and that RANKL signaling small dynamic concentration defects are greatly amplified by the continuous alternation of absorption and formation resulting in large structural bone defects. This work contributes to the computational modeling of complex systems with a multilevel approach connecting formal languages and agent-based simulation tools.

In this paper we investigate plant photosynthesis and microbial fuel cells. We report the following: 1) we introduce and validate a novel multi-objective optimization algorithm, PMO2; 2) in photosynthesis we increase the yield of 135%, while in Geobacter sulfurreducens we determine the tradeoff for growth versus redox properties; 3) finally, we discuss Pareto-Front as an estimator of robust metabolic pathways.

In Delay Tolerant Networks (DTN), as disconnections between nodes are frequent, establishing the routing paths from the source node to the destination node may not be possible. Messages are forwarded in DTN similarly to an infective disease spreading among humans. This paper sets up an evolving model of messages delivery. Also, in this paper, we propose the 'crash' issue. Because large part of persons will turn off their mobile phones and discard the messages received, the 'crash' could happen in DTNs. The 'crash' in DTN means large part of the copies of messages are discarded within a short time duration. The evolving model is validated by a message forwarding simulation among mobile nodes. In the simulation, we study the influence of 'crashes' on the delivery ratio and delay of Epidemic routing, SprayAndWait and Prophet routing protocols.

In previous work we presented supportive computational analysis on newly found biological evidence which indicates the existence of a dormant Hematopoietic Stem Cell (HSC) population. Through extensive modelling of experimental DNA label-retaining cell data we showed that an ordinary differential equation (ODE) model can successfully capture a heterogeneous HSC population structure. The ODE model's analytical tractability made it especially suitable for parameter estimation in contrast to an earlier agent-based model we developed on a related but independent Bromodeoxyuridine (BrdU) dataset. For this current study we explore the predictive power of the same agent-based model on the larger more elaborate BrdU dataset by comparing the BrdU detection threshold (BDT) estimates of both the ODE (continuous BDT implementation) and agent-based (discrete BDT implementation) models. We therefore re-estimate the HSC division parameters using the agent-based model which entailed a brute-force search approach. In order to cover a worthwhile region of the parameter hyperspace within a reasonable amount of time we executed our search algorithm in parallel over the EGEE Grid environment. Our results indicate that the agent-based model more or less supports the same conclusions as the ODE model. However, actual cell division rate estimates as well as model prediction differ slightly for the same set of parameters. The estimates for the dormant HSC proportion and BDT, on the other hand, are in strong agreement with the ODE estimates. For the BDT in particular, this is an encouraging result as the two models have a very different approach in their implementation of the BDT.

We have studied the $\mathbf{C_3}$ photosynthetic carbon metabolism centering our investigation on the following four design principles. (1) Optimization of the photosynthetic rate by modifying the partitioning of resources between the different enzymes of the $\mathbf{C_3}$ photosynthetic carbon metabolism using a constant amount of protein-nitrogen. (2) Identify sensitive and less sensitive enzymes of the studied model. (3) Maximize photosynthetic productivity rate through the choice of robust enzyme concentrations using a new precise definition of robustness. (4) Modeling photosynthetic carbon metabolism as a multi-objective problem of two competing biological selection pressures: light-saturated photosynthetic rate versus total protein-nitrogen requirement. Using the designed single-objective optimization algorithms, PAO and A-CMA-ES, we have obtained an increase in photosynthetic productivity of the $\mathbf{135\%}$ from 15.486 $\mathbf{\mu mol~m^{-2}s^{-1}}$ to $\mathbf{36.382~\mu mol~m ^{-2}s^{-1}}$, and improving the previous best-found photosynthetic productivity value ($\mathbf{27.261}$ $\mathbf{\mu mol~m ^{-2}s^{-1}}$, $\mathbf{76\%}$ of enhancement). Optimized enzyme concentrations express a maximal local robustness ($\mathbf{100\%}$) and a high global robustness ($\mathbf{97.2\%}$), satisfactory properties for a possible ``in vitro'' manufacturing of the optimized pathway. Morris sensitivity analysis shows that 11 enzymes over 23 are high sensitive enzymes, i.e., the most influential enzymes of the carbon metabolism model. Finally, we have obtained the trade-off between the maximization of the leaf $\mathbf{CO_2}$ uptake rate and the minimization of the total protein-nitrogen concentration. This trade-off search has been carried out for the three $\mathbf{c_i}$ concentrations referring to the estimate of $\mathbf{CO_2}$ concentration in the atmosphere characteristic of 25 million years ago, nowadays and in 2100 a.C. Remarkably, the three Pareto frontiers identify the highest photosynthetic productivity rates together with the fewest protein-nitrogen usage.

DNA motif discovery is an important problem for deciphering gene regulation. Motifs usually contain gaps (spaced) and are more complex than contiguously conserved (monad) patterns. Existing algorithms mostly address monad motifs, and methods for spaced motifs impose various constraints on gaps, which may affect the discovery of complex motifs. In this paper, we propose Genetic Algorithm (GA) for Spaced Motifs Elicitation on Nucleotides (GASMEN), which searches from a wide range of possible widths (4-25) and relaxes substantial constraints. GASMEN employs submotif indexing to partition the search space into smaller sub-space for GA to easier reach optimality. Multiple-motif control is employed and probabilistic refinements are proposed to improve motif quality respectively. The preliminary results on real spaced motifs demonstrate that GASMEN is promising to find more accurate motifs and optimal widths, compared with the state-of-the-art method, SPACE. GASMEN is also capable of finding monad motifs, outperforming both Weeder and SPACE on most of the 8 real datasets.

BACKGROUND: Several diseases, many of which nowadays pandemic, consist of multifactorial pathologies. Paradigmatic examples come from the immune response to pathogens, in which cases the effects of different infections combine together, yielding complex mutual feedback, often a positive one that boosts infection progression in a scenario that can easily become lethal. HIV is one such infection, which weakens the immune system favouring the insurgence of opportunistic infections, amongst which Tuberculosis (TB). The treatment with antiretroviral therapies has shown effective in reducing mortality. An in-depth understanding of complex systems, like the one consisting of HIV, TB and related therapies, is an open great challenge, on the boundaries of bioinformatics, computational and systems biology. RESULTS: We present a simplified formalisation of the highly dynamic system consisting of HIV, TB and related therapies, at the cellular level. The progression of the disease (AIDS) depends hence on interactions between viruses, cells, chemokines, the high mutation rate of viruses, the immune response of individuals and the interaction between drugs and infection dynamics. We first discuss a deterministic model of dual infection (HIV and TB) which is able to capture the long-term dynamics of CD4 T cells, viruses and Tumour Necrosis Factor (TNF). We contrast this model with a stochastic approach which captures intrinsic fluctuations of the biological processes. Furthermore, we also integrate automated reasoning techniques, i.e. probabilistic model checking, in our formal analysis. Beyond numerical simulations, model checking allows general properties (effectiveness of anti-HIV therapies) to be verified against the models by means of an automated procedure. Our work stresses the growing importance and flexibility of model checking techniques in bioinformatics. In this paper we i) describe HIV as a complex case of infectious diseases; ii) provide a number of different formal descriptions that suitably account for aspects of interests; iii) suggest that the integration of different models together with automated reasoning techniques can improve the understanding of infections and therapies through formal analysis methodologies. CONCLUSION: We argue that the described methodology suitably supports the study of viral infections in a formal, automated and expressive manner. We envisage a long-term contribution of this kind of approaches to clinical Bioinformatics and Translational Medicine.

BACKGROUND: In the diagnosis of complex diseases such as neurological pathologies, a wealth of clinical and molecular information is often available to help the interpretation. Yet, the pieces of information are usually considered in isolation and rarely integrated due to the lack of a sound statistical framework. This lack of integration results in the loss of valuable information about how disease associated factors act synergistically to cause the complex phenotype. RESULTS: Here, we investigated complex psychiatric diseases as networks. The networks were used to integrate data originating from different profiling platforms. The weighted links in these networks capture the association between the analyzed factors and allow the quantification of their relevance for the pathology. The heterogeneity of the patient population was analyzed by clustering and graph theoretical procedures. We provided an estimate of the heterogeneity of the population of schizophrenia and detected a subgroup of patients featuring remarkable abnormalities in a network of serum primary fatty acid amides. We compared the stability of this molecular network in an extended dataset between schizophrenia and affective disorder patients and found more stable structures in the latter. CONCLUSION: We quantified robust associations between analytes measured with different profiling platforms as networks. The methodology allows the quantitative evaluation of the complexity of the disease. The identified disease patterns can then be further investigated with regards to their diagnostic utility or help in the prediction of novel therapeutic targets. The applied framework is able to enhance the understanding of complex psychiatric diseases, and may give novel insights into drug development and personalized medicine approaches.

Random projection (RP) is a common technique for dimensionality reduction under L2 norm for which many significant space embedding results have been demonstrated. However, many similarity search applications often require very low dimension embeddings in order to reduce overhead and boost performance. Inspired by the use of symmetric probability distributions in previous work, we propose a novel RP algorithm, Beta Random Projection, and give its probabilistic analyses based on Beta and Gaussian approximations. We evaluate the algorithm in terms of standard similarity metrics with other RP algorithms as well as the singular value decomposition (SVD). Our experimental results show that BRP preserves both similarity metrics well and, under various dataset types including random point sets, text (TREC5) and images, provides sharper and consistent performance.

The study of metabolic pathways is becoming increasingly important to exploit an integrated, systems-level approach for optimizing a desired cellular property or phenotype. In this context, the integration of genomics data with genetic, metabolic and regulatory models is essential because the systematic design of artificial, biological systems requires the identification of robust building blocks like gene promoters, metabolic pathways or genetic circuits taken from natural organisms, and manipulated to develop ad hoc features. Computational tools allowing precise descriptions of natural pathways might thus allow improving the performance of artificial routes. In this review, we introduce the most recent bioinformatics tools enabling detailed characterizations of metabolic pathways in bacteria from different perspectives.

BACKGROUND: Phylogenetic methods are well-established bioinformatic tools for sequence analysis, allowing to describe the non-independencies of sequences because of their common ancestor. However, the evolutionary profiles of bacterial genes are often complicated by hidden paralogy and extensive and/or (multiple) horizontal gene transfer (HGT) events which make bifurcating trees often inappropriate. In this context, plasmid sequences are paradigms of network-like relationships characterizing the evolution of prokaryotes. Actually, they can be transferred among different organisms allowing the dissemination of novel functions, thus playing a pivotal role in prokaryotic evolution. However, the study of their evolutionary dynamics is complicated by the absence of universally shared genes, a prerequisite for phylogenetic analyses. RESULTS: To overcome such limitations we developed a bioinformatic package, named Blast2Network (B2N), allowing the automatic phylogenetic profiling and the visualization of homology relationships in a large number of plasmid sequences. The software was applied to the study of 47 completely sequenced plasmids coming from Escherichia, Salmonella and Shigella spps. CONCLUSION: The tools implemented by B2N allow to describe and visualize in a new way some of the evolutionary features of plasmid molecules of Enterobacteriaceae; in particular it helped to shed some light on the complex history of Escherichia, Salmonella and Shigella plasmids and to focus on possible roles of unannotated proteins.The proposed methodology is general enough to be used for comparative genomic analyses of bacteria.

Recent analyses of human genome sequences have given rise to impressive advances in identifying non-synonymous single nucleotide polymorphisms (nsSNPs). By contrast, the annotation of nsSNPs and their links to diseases are progressing at a much slower pace. Many of the current approaches to analysing disease-associated nsSNPs use primarily sequence and evolutionary information, while structural information is relatively less exploited. In order to explore the potential of such information, we developed a structure-based approach, Bongo (Bonds ON Graph), to predict structural effects of nsSNPs. Bongo considers protein structures as residue-residue interaction networks and applies graph theoretical measures to identify the residues that are critical for maintaining structural stability by assessing the consequences on the interaction network of single point mutations. Our results show that Bongo is able to identify mutations that cause both local and global structural effects, with a remarkably low false positive rate. Application of the Bongo method to the prediction of 506 disease-associated nsSNPs resulted in a performance (positive predictive value, PPV, 78.5%) similar to that of PolyPhen (PPV, 77.2%) and PANTHER (PPV, 72.2%). As the Bongo method is solely structure-based, our results indicate that the structural changes resulting from nsSNPs are closely associated to their pathological consequences.

The field of forensic science is increasingly based on biomolecular data and many European countries are establishing forensic databases to store DNA profiles of crime scenes of known offenders and apply DNA testing. The field is boosted by statistical and technological advances such as DNA microarray sequencing, TFT biosensors, machine learning algorithms, in particular Bayesian networks, which provide an effective way of evidence organization and inference. The aim of this article is to discuss the state of art potentialities of bioinformatics in forensic DNA science. We also discuss how bioinformatics will address issues related to privacy rights such as those raised from large scale integration of crime, public health and population genetic susceptibility-to-diseases databases.

UNLABELLED: We describe MotifScorer, a program for systematic genome-wide identification of transcription sites. The program uses a compendium of gene expression microarrays and implements state-of-art partial least squares (PLSs) based regression and stepwise regression procedures. Candidate motifs from the upstream sequences of groups of co-regulated genes are identified and assigned a score using genomic background models and available motif finding tools. The use of a large library of expression data allows statistical comparative analysis of the specificity of motifs identified in different conditions. AVAILABILITY: MotifScorer, which is written in Java and Matlab, manual and example files are available from the authors.

The identification of cell cycle-related genes is still a difficult task, even for organisms with relatively few genes such as the fission yeast. Several gene expression studies have been published on S. pombe showing similarities but also discrepancies in their results. We introduce a network in which the weight of each link is a function of the phase difference between the expression peaks of two genes. The analysis of the stability of the clustering through the computation of an entropy parameter reveals a structure made of four clusters, the first one corresponding to a robustly connected M-G1 component, the second to genes in the S phase, and the third and fourth to two G2 components. They are separated by bottleneck structures that appear to correspond to cell cycle checkpoints. We identify a number of genes that are located on these bottlenecks. They represent a novel group of cell cycle regulatory genes. They all show interesting functions, and they are supposed to be involved in the regulation of the transition from one phase to the next. We therefore present a comparison of the available studies on the fission yeast cell cycle and a general statistical bioinformatics methodology to find bottlenecks and gene community structures based on recent developments in network theory.

SUMMARY: We describe an algorithm and software tool for comparing alternative phylogenetic trees. The main application of the software is to compare phylogenies obtained using different phylogenetic methods for some fixed set of species or obtained using different gene sequences from those species. The algorithm pairs up each branch in one phylogeny with a matching branch in the second phylogeny and finds the optimum 1-to-1 map between branches in the two trees in terms of a topological score. The software enables the user to explore the corresponding mapping between the phylogenies interactively, and clearly highlights those parts of the trees that differ, both in terms of topology and branch length. AVAILABILITY: The software is implemented as a Java applet at http://www.mrc-bsu.cam.ac.uk/personal/thomas/phylo_comparison/comparison_page.html. It is also available on request from the authors.

The authors presented a novel approach for keyword searching for file sharing applications based on a simple hash function and extension to current distributed hash tables (DHT) topology. Depart from standard hash methods on DHT systems, this approach is to develop a locality preserving hash function so that objects containing the same keyword tend to have similar hash values under Hamming metric. It is shown that this hash function is locality preserving and proposes the embedding of keyword edges to facilitate query processing. As such, several class of problems in keyword search were reduced into a network multicast problem.

MOTIVATION: Understanding the mechanisms that determine gene expression regulation is an important and challenging problem. A common approach consists of identifying DNA-binding sites from a collection of co-regulated genes and their nearby non-coding DNA sequences. Here, we consider a regression model that linearly relates gene expression levels to a sequence matching score of nucleotide patterns. We use Bayesian models and stochastic search techniques to select transcription factor binding site candidates, as an alternative to stepwise regression procedures used by other investigators. RESULTS: We demonstrate through simulated data the improved performance of the Bayesian variable selection method compared to the stepwise procedure. We then analyze and discuss the results from experiments involving well-studied pathways of Saccharomyces cerevisiae and Schizosaccharomyces pombe. We identify regulatory motifs known to be related to the experimental conditions considered. Some of our selected motifs are also in agreement with recent findings by other researchers. In addition, our results include novel motifs that constitute promising sets for further assessment. AVAILABILITY: The Matlab code for implementing the Bayesian variable selection method may be obtained from the corresponding author.

Genome studies have become central to a wide range of biological areas. This paper is intended to discuss the state of art and potentiality of using statistics in genome comparison. The problem of dependence and dimensionality in genome data is addressed first. The focus is on using phylogenetic methods in genome comparison, on combining sequence and gene expression data to find transcription sites, and on multiple hypothesis testing in gene expression data. The paper concludes with speculations on the future course of statistical and bioinformatics analyses.

MOTIVATION: At a recent meeting, the wavelet transform was depicted as a small child kicking back at its father, the Fourier transform. Wavelets are more efficient and faster than Fourier methods in capturing the essence of data. Nowadays there is a growing interest in using wavelets in the analysis of biological sequences and molecular biology-related signals. RESULTS: This review is intended to summarize the potential of state of the art wavelets, and in particular wavelet statistical methodology, in different areas of molecular biology: genome sequence, protein structure and microarray data analysis. I conclude by discussing the use of wavelets in modeling biological structures.

Abstract Motivation: There is a growing literature on wavelet theory and wavelet methods showing improvements on more classical techniques, especially in the contexts of smoothing and extraction of fundamental components of signals. G+Cpatterns occur at different lengths (scales) and, for this reason, G+Cplots are usually difficult to interpret. Current methods for genome analysis choose a window size and compute a \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \({\chi}^{2}\) \end{document}statistics of the average value for each window with respect to the whole genome. Results: Firstly, wavelets are used to smooth G+Cprofiles to locate characteristic patterns in genome sequences. The method we use is based on performing a \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \({\chi}^{2}\) \end{document}statistics on the wavelet coefficients of a profile; thus we do not need to choose a fixed window size, in that the smoothing occurs at a set of different scales. Secondly, a wavelet scalogram is used as a measure for sequence profile comparison; this tool is very general and can be applied to other sequence profiles commonly used in genome analysis. We show applications to the analysis of Deinococcus radiodurans chromosome I, of two strains of Helicobacter pylori (26695, J99) and two of Neisseria meningitidis (serogroup B strain MC58 and serogroup A strain Z2491). We report a list of loci that have different G+Ccontent with respect to the nearby regions; the analysis of N. meningitidis serogroup B shows two new large regions with low G+Ccontent that are putative pathogenicity islands. Availability: Software and numerical results (profiles, scalograms, high and low frequency components) for all the genome sequences analyzed are available upon request from the authors. Contact: [email protected] To whom correspondence should be addressed.

MOTIVATION: A non-parametric method, based on a wavelet data-dependent threshold technique for change-point analysis, is applied to predict location and topology of helices in transmembrane proteins. A new propensity scale generated from a transmembrane helix database is proposed. RESULTS: We show that wavelet change-point performs well for smoothing hydropathy and transmembrane profiles generated using different scales. We investigate which wavelet bases and threshold functions are overall most appropriate to detect transmembrane segments. Prediction accuracy is based on the analysis of two data sets used as standard benchmarks for transmembrane prediction algorithms. The analysis of a test set of 83 proteins results in accuracy per segment equal to 98.2%; the analysis of a 48 proteins blind-test set, i.e. containing proteins not used to generate the propensity scales, results in accuracy per segment equal to 97.4%. We believe that this method can also be applied to the detection of boundaries of other patterns such as G + Cisochores and dot-plots. AVAILABILITY: The transmembrane database, TMALN and source code are available upon request from the authors.