Integrating external medical knowledge into longitudinal electronic health record modeling is a prevailing paradigm to mitigate clinical data sparsity.However, existing approaches face a reliability-timeliness dilemma, struggling to balance the structural authority of static ontologies with the reasoning flexibility of large language models.Furthermore, most frameworks overlook the risk of relative negative transfer, where indiscriminately fusing task-irrelevant knowledge can introduce noise or even cause conflicts that weakens patient-specific signals.In this paper, we propose TrustKE, a Trustworthy Knowledge Enhancement framework.First, we construct a dual-layer knowledge graph that anchors dynamic, evidence-based chain-of-thought reasoning from medical literature within the stable structure of medical knowledge graph.Second, we introduce a task-adaptive knowledge selection mechanism that dynamically optimizes the graph, retaining only task-specific signals.Extensive experiments on MIMIC-III and MIMIC-IV across four clinical tasks show that TrustKE outperforms state-of-the-art baselines.Our analysis confirms that TrustKE effectively mitigates negative transfer while offering transparent reasoning for clinical decision-making.

Drug recommendation based on electronic health record (EHR) is fundamental to effective disease treatment. Similar to commercial sequence-based recommendation systems, the accuracy of drug recommendation largely depends on precise patient modeling. However, patient modeling is more complex, as it not only requires sequence modeling of patient's disease course, but also needs to refer to the information of patients with similar medical medication. In EHR data, many patients have only one visit record, and the similarity between patients is often vague and unclear, which may cause noise and ambiguity. This leads to significant challenges for the drug recommendation field, especially when patient records are sparse or when patient similarity is vague. To address the above challenges, we propose RaVSNet (Relevance aware Visit Similarity Network), which improves drug recommendation by leveraging both longitudinal and transversal visit similarity and integrating medical relevance knowledge. RaVSNet utilizes multi-dimensional visit information similar to the patient's current visit as a reference, and employs a relevance-aware network to explicitly model the matching relationships between medical conditions and medications. Additionally, RaVSNet designs a general pretraining framework specifically for drug recommendation, including two tasks, Medication Sequence Reconstruction (MSR) and Causal Effect Inference (CEI), to discover the deep connections between medical information and medications. Experimental results on two public EHR datasets, MIMIC-III and MIMIC-IV demonstrate that the proposed algorithm outperforms state-of-the-art methods, yielding more accurate drug recommendation combinations, and the proposed general pretraining framework can be seamlessly integrated into most drug recommendation methods to achieve performance improvements.

The effective fusion of pathological images and molecular omics holds significant potential for precision medicine. However, pathological and molecular data are highly heterogeneous, and large-scale multi-modal cancer data often suffer from incomplete information. Predicting clinical tasks from such imbalanced multi-modal data presents a major challenge. Therefore, we propose a confidence-aware adaptive fusion framework CAFusion. The framework adopts a modular design, providing independent and flexible modal feature learning modules to capture high-quality features. To address issues of modal imbalance caused by heterogeneous and incomplete modal, we design a confidence-aware method that evaluates the features of each modal and automatically adjusts their weights. To effectively fuse pathological and molecular modals, we propose an adaptive deep network, which features a flexible, non-fixed layer structure that effectively extracts hidden joint information from multi-modal features, ensuring high generalizability. Experiment results demonstrate that the performance of the CAFusion framework outperforms other state-of-the-art methods, both on complete and incomplete datasets. Moreover, the CAFusion framework offers reasonable medical interpretability.

Single-vector dense retrieval models, which are foundational to modern Retrieval-Augmented Generation (RAG) systems, struggle to represent the multifaceted semantics of complex documents, particularly in specialized fields like biomedicine. This semantic bottleneck limits their ability to provide comprehensive context for generation tasks. To address this, we propose ELK-Multi, a novel multi-vector retrieval framework that constructs fine-grained document representations through knowledge-enhanced entity-type clustering. By leveraging a knowledge-aware encoder, ELK-Multi first identifies and groups entities by their type, generating a distinct vector for each semantic cluster. These targeted representations are then combined with a global document vector using principled aggregation strategies to balance fine-grained detail with holistic context. Extensive experiments on the TREC-COVID and NFCorpus datasets validate our approach, where ELK-Multi establishes new state-of-the-art results in NDCG and Recall. This is complemented by a detailed efficiency analysis demonstrating that our model achieves this performance while remaining within the efficient dual-encoder paradigm, alongside a qualitative analysis with case studies and visualizations.

Bone marrow blood cell morphology examination is a critical diagnostic tool for blood diseases such as leukemia. Accurate identification of the types and proportions of bone marrow blood cells plays a pivotal role in achieving precise diagnoses. However, due to the diversity and structural similarities among certain types of bone marrow cells, existing classification models often struggle with accurate discrimination, leading to a higher likelihood of misclassification. To address these challenges, this paper introduces BMCNet-A, a classification model specifically designed for bone marrow cells. The model incorporates multi-scale structural features and an integrated attention mechanism to enhance classification performance. By leveraging critical scale information of cells and capturing contextual relationships between the nucleus and cytoplasm within the images, BMCNet-A effectively reduces the misclassification rate. Experimental results demonstrate that BMCNet-A achieves state-of-the-art performance on the BMC_GD dataset, with an Accuracy of 90.83% and an F1-Score of 90.56%, representing improvements of 1.02 and 1.40 percentage points, respectively, over ConvNeXt-T. Furthermore, the model exhibits strong generalization capabilities, achieving an Accuracy of 89.81% and an F1-Score of 83.65% on the BMCC dataset, outperforming ConvNeXt-T by 0.64 and 2.08 percentage points, respectively.

In recent years, the incidence of breast cancer has risen rapidly, making it one of the leading types of cancer worldwide. Mammography is the primary method for early breast cancer screening and can significantly improve the survival rate of patients. Breast mass detection is a critical step in breast cancer diagnosis. Radiologists typically integrate mammographic images from different views to identify masses and further determine their malignancy or benignity. Despite progress in existing computer-aided diagnosis methods, the inability to fully utilize the correlation information between multiple views has hindered the accuracy and interpretability of diagnoses. To address this issue, we propose MVBMR, which encompasses two critical stages: detection and classification. In the detection stage, we design and integrate WCA-RCNN and MV-RCNN. WCA-RCNN focuses on fusing information from bilateral mammograms to identify highly concealed masses, while MV-RCNN emphasizes combining information from ipsilateral mammograms to reduce false positives. In the classification stage, we employ multi-task learning to integrate mass shape and edge features, supporting benign and malignant classification with morphological information. Experimental results on the public mammography dataset DDSM demonstrate that the proposed detection method achieves higher Recall values at FPPI levels of 1.0 and 2.0 compared to the existing state-of-the-art methods. Additionally, the multi-task design improves the mass classification accuracy by approximately 4%, fully validating its effectiveness and superiority.

Deep graph learning models have recently been developed to learn from various graphs that are prevalent in describing and modeling complex systems, including those in bioinformatics. However, a versatile explanation method for uncovering the general graph patterns that guide deep graph models in making predictions remains elusive. In this paper, we propose DGX, a novel deep graph model explainer that generates explanatory graphs to explain trained, opaque-box deep graph models. Its effectiveness is demonstrated by producing multiple graphs that collectively encode the structural knowledge captured by the graph neural network on both synthetic and real graph data. Importantly, DGX can produce diverse explanations by generating a set of distinguishable graphs and can provide customized explanations based on prior knowledge or constraints specified by users. We apply DGX to explain a mutagenicity prediction model by exploring the underlying groups of mutagenic compounds, and we explain the model on brain functional networks by revealing the structural patterns that enable the model to differentiate autism spectrum disorder from healthy controls. These findings offer an effective, diverse, and customized approach to explaining the underlying mechanisms and enhancing the understanding of models learned from real graph data, particularly in fields such as biomedicine and bioinformatics.

Traffic accidents remain a significant concern due to their potential severity and impact on society. The rapid and accurate detection of car damage is increasingly crucial. Manual assessment of car damage usually relies on multi-view car images taken at the scene, which can provide richer information for damage assessment. However, most car damage algorithms are based on single-view datasets, and it is hard to fully leverage the complementary information and alignment information between distant view and close-up images. In this paper, we propose the Multi-View Car Damage Detection model (MVA-CDD), comprising three key modules: Feature Split (FS), Feature Fusion (FF), and Image Alignment (IA). The FS module extracts global and detailed information from distant-view and close-up images separately, which are then combined by the FF module. The IA module effectively aligns car damage information in distant-view and close-up images to correct errors and biases. Meanwhile, we created the new Car Damage Detection Multi-view dataset (CDDM), which has a significant advantage in both image quantity and diversity across categories, addressing the shortcomings of existing multi-view datasets. Our proposed MVA-CDD outperforms the state-of-the-art single-view and multi-view models with the dataset. Results from ablation studies further confirm the efficiency of MVA-CDD. This study contributes to optimizing the car damage detection and claims adjudication process, leading to significant labor and material cost savings. CDDM dataset is available athttps://github.com/SCUT-CCNL/CDDM.

In recent years, medical image segmentation technology has made great progress. However, the annotated data of 3D medical images is relatively small, and although fewshot segmentation can solve this problem, there are still many challenges. Too few samples of support images may lead to the fact that it is difficult to fully represent 3D medical images, especially the important 3D spatial information, and the global correlations between support and query images are not fully utilized. In this paper, we propose a novel few-shot 3D medical image segmentation pipeline framework, SMFT-Net, which can efficiently accomplish the 3D medical image segmentation task using only one labeled sample. Specifically, we proposed pretrained feature transfer module (PFTM) and bidirectional feature transfer module (BFTM) for multiple feature transfer of 3D medical image. PFTM can be used for 3D feature transfer to ensure that the 3D spatial information of medical images is preserved. And BFTM can perform bi-directional feature transfer between the query image and the support image to eliminate extraneous information from the surrounding pixels. Extensive experiments on four medical image datasets demonstrate that our method outperforms the state-of-the-art methods.

When applied across different scenes, hyperspectral image (HSI) classification models often struggle to generalize due to the data distribution disparities and labels’ scarcity, leading to domain shift (DS) problems. Recently, the high-level semantics from text has demonstrated the potential to address the DS problem, by improving the generalization capability of image encoders through aligning image-text pairs. However, the main challenge still lies in crafting appropriate texts that accurately represent the intricate interrelationships and the fragmented nature of land cover in HSIs and effectively extracting spectral-spatial features from HSI data. This article proposes a domain generalization (DG) method, EHSnet, to address these issues by leveraging multilayered explicit high-level semantic (EHS) information from different types of texts to provide precisely relevant semantic information for the image encoder. A multilayered EHS information paradigm is well-defined, aiming to extract the HSI’s intricate interrelationships and the fragmented land-cover features, and a dual-residual encoder connected by a 2-D convolution is designed, which combines CNNs with residual structure and Vision Transformers (ViTs) with short-range cross-layer connections to explore the spectral-spatial features of HSIs. By aligning text features with image features in the semantic space, EHSnet improves the representation capability of the image encoder and is endowed with zero-shot generalization ability for cross-scene tasks. Extensive experiments conducted on three hyperspectral datasets, including Houston, Pavia, and XS datasets, validate the effectiveness and superiority of EHSnet, with the Kappa coefficient improved by 8.17%, 3.22%, and 3.62% across three datasets compared to the state-of-the-art (SOTA) methods. The code is available athttps://github.com/SCUT-CCNL/EHSnet.

The performance of biomedical information retrieval greatly depends on biomedical knowledge; however the knowledge of available medical knowledge base is often incomplete and out-of-dated. To solve the problem that incomplete knowledge bases cannot provide the medical knowledge required for biomedical information retrieval, the paper proposes an Entity Relation Aware Graph Neural Ranking model (ERAGNR), aiming to fully leverage the internal knowledge of the document to alleviate the problem caused by incomplete external knowledge bases. ERAGNR mines the relationships between biomedical entities in the document through entity relation extraction and combines them with external knowledge. It increases the semantic association and reduces the semantic gap between the query and the document. The method first constructs a knowledge-query graph and a document-entity graph, and then fuses the two graphs to obtain a knowledge-query-document-entity graph. In a multi-task learning framework that combines text retrieval and relation extraction tasks, ERAGNR employs a shared text encoder and a graph neural network. This enables ERAGNR to learn semantic matching patterns between queries and documents and recognize relationships between entities in the documents. As a result, the model can capture semantic matching signals between entity relationships in the context and queries. The experimental results show that ERAGNR outperforms the state-of-the-art models. Through biomedical relation extraction task, the model can learn the ability to capture the context of the entity relations in the document, so that the model can more accurately match the semantics between the query and the document.

Advanced deep learning methods, especially graph neural networks (GNNs), are increasingly expected to learn from brain functional network data and predict brain disorders. In this paper, we proposed a novel Transformer and snowball encoding networks (TSEN) for brain functional network classification, which introduced Transformer architecture with graph snowball connection into GNNs for learning whole-graph representation. TSEN combined graph snowball connection with graph Transformer by snowball encoding layers, which enhanced the power to capture multi-scale information and global patterns of brain functional networks. TSEN also introduced snowball graph convolution as position embedding in Transformer structure, which was a simple yet effective method for capturing local patterns naturally. We evaluated the proposed model by two large-scale brain functional network datasets from autism spectrum disorder and major depressive disorder respectively, and the results demonstrated that TSEN outperformed the state-of-the-art GNN models and the graph-transformer based GNN models.

Deep learning methods have been applied for dynamic brain functional network analysis recently. However, they are usually restricted by the complex spatio-temporal dynamics and the limited labeled data. In this paper, we proposed a stacked Transformer neural network, namely BrainPST, to capture spatio-temporal patterns for dynamic brain functional network classification. BrainPST model integrated spatial and temporal information by stacking two Transformers: one for learning snapshot networks and the other for learning sequence of functional connections. Unlike recent models, BrainPST was designed to pre-train the stacked Transformer network by leveraging unlabeled existing brain imaging data. A pre-training framework with designed pre-training strategies was proposed to learn general spatio-temporal representations from the unlabeled brain networks, and to fine-tune the pre-trained model in downstream tasks. BrainPST is a conceptually simple and effective model. The results of experiments showed the BrainPST model without pre-training achieved comparative performance with the recent models, and the pre-trained BrainPST obtained new state-of-the-art performance. The pre-trained BrainPST improved 3.92% of AUC compared with the model without pre-training.

The morphological examination of bone marrow (BM) cells is an essential basis for the diagnosis and assessment of hematologic diseases. However, manual recognition of cell class and quantification in microscopic images of BM cells is time-consuming, labor-intensive, and prone to errors. Therefore, it is crucial to study an efficient algorithm for the automatic recognition of BM cells. Clinically, to determine the BM cell classes, experienced examiners usually rely on surrounding cells as references to provide auxiliary confirmation, but there is no model specially designed for contextual relationships of BM cells. To extract cell features comprehensively and explore the contextual relationships among BM cells, we propose RC RCNN based on Mask R-CNN, which includes positional encoding (PE), ROI self-attention module (RSAM), and ROI channel attention module (RCAM) as an extension to Mask RCNN. The PE and RCAM enable better capturing of contextual information between BM cells. By using the RSAM, the model directs its attention toward internal cell features while effectively suppressing irrelevant background information. Experimental results demonstrate the outstanding performance of the proposed method, with the AP and F1-Score reaching 73.21% and 87.71%, respectively, an improvement of about 2.3% and 2.2% over the baseline.

The development of omics data and biomedical images has greatly advanced the progress of precision medicine in diagnosis, treatment, and prognosis. The fusion of omics and imaging data, i.e., omics-imaging fusion, offers a new strategy for understanding complex diseases. However, due to a variety of issues such as the limited number of samples, high dimensionality of features, and heterogeneity of different data types, efficiently learning complementary or associated discriminative fusion information from omics and imaging data remains a challenge. Recently, numerous machine learning methods have been proposed to alleviate these problems. In this review, from the perspective of fusion levels and fusion methods, we first provide an overview of preprocessing and feature extraction methods for omics and imaging data, and comprehensively analyze and summarize the basic forms and variations of commonly used and newly emerging fusion methods, along with their advantages, disadvantages and the applicable scope. We then describe public datasets and compare experimental results of various fusion methods on the ADNI and TCGA datasets. Finally, we discuss future prospects and highlight remaining challenges in the field.

The traditional vehicle damage assessment process is complicated and time-consuming, asking for intelligent methods for detecting vehicle damage. At present, most damage detection methods for vehicles require two models to detect damage and component where the damage is located separately, which is complex and inefficient. To this end, we propose an end-to-end multi-detection model named MD R-CNN, which simultaneously outputs damage detection and component recognition results by adding an extra classification branch. To improve the positioning precision of detection, the regression of the detection box adopts a self-attention convolution head (SA-Head) composed of a residual module and two SC Attention modules; Moreover, since there are few damage annotation datasets available, D-FPN is proposed to enhance the multi-scale detection performance. The experimental results show that MD R-CNN increases the Average precision (AP) by about 2.4% on vehicle damage datasets, and the precision can reach 80.22%, which has a favorable performance.

BACKGROUND: Autism spectrum disorders (ASD) imply a spectrum of symptoms rather than a single phenotype. ASD could affect brain connectivity at different degree based on the severity of the symptom. Given their excellent learning capability, graph neural networks (GNN) methods have recently been used to uncover functional connectivity patterns and biological mechanisms in neuropsychiatric disorders, such as ASD. However, there remain challenges to develop an accurate GNN learning model and understand how specific decisions of these graph models are made in brain network analysis. RESULTS: In this paper, we propose a graph attention network based learning and interpreting method, namely GAT-LI, which learns to classify functional brain networks of ASD individuals versus healthy controls (HC), and interprets the learned graph model with feature importance. Specifically, GAT-LI includes a graph learning stage and an interpreting stage. First, in the graph learning stage, a new graph attention network model, namely GAT2, uses graph attention layers to learn the node representation, and a novel attention pooling layer to obtain the graph representation for functional brain network classification. We experimentally compared GAT2 model's performance on the ABIDE I database from 1035 subjects against the classification performances of other well-known models, and the results showed that the GAT2 model achieved the best classification performance. We experimentally compared the influence of different construction methods of brain networks in GAT2 model. We also used a larger synthetic graph dataset with 4000 samples to validate the utility and power of GAT2 model. Second, in the interpreting stage, we used GNNExplainer to interpret learned GAT2 model with feature importance. We experimentally compared GNNExplainer with two well-known interpretation methods including Saliency Map and DeepLIFT to interpret the learned model, and the results showed GNNExplainer achieved the best interpretation performance. We further used the interpretation method to identify the features that contributed most in classifying ASD versus HC. CONCLUSION: We propose a two-stage learning and interpreting method GAT-LI to classify functional brain networks and interpret the feature importance in the graph model. The method should also be useful in the classification and interpretation tasks for graph data from other biomedical scenarios.

The curse of dimensionality, which is caused by high-dimensionality and low-sample-size, is a major challenge in gene expression data analysis. However, the real situation is even worse: labelling data is laborious and time-consuming, so only a small part of the limited samples will be labelled. Having such few labelled samples further increases the difficulty of training deep learning models. Interpretability is an important requirement in biomedicine. Many existing deep learning methods are trying to provide interpretability, but rarely apply to gene expression data. Recent semi-supervised graph convolution network methods try to address these problems by smoothing the label information over a graph. However, to the best of our knowledge, these methods only utilize graphs in either the feature space or sample space, which restrict their performance. We propose a transductive semi-supervised representation learning method called a hierarchical graph convolution network (HiGCN) to aggregate the information of gene expression data in both feature and sample spaces. HiGCN first utilizes external knowledge to construct a feature graph and a similarity kernel to construct a sample graph. Then, two spatial-based GCNs are used to aggregate information on these graphs. To validate the model's performance, synthetic and real datasets are provided to lend empirical support. Compared with two recent models and three traditional models, HiGCN learns better representations of gene expression data, and these representations improve the performance of downstream tasks, especially when the model is trained on a few labelled samples. Important features can be extracted from our model to provide reliable interpretability.

Graph neural networks have been developed for various node classification tasks in graph data. However, the noise in graph data would affect the effectiveness of training and prediction of these graph learning models. In this paper, we propose a graph learning approach, named GCN_CL, which introduces a confident learning method into the graph convolutional networks model to classify nodes in graph data with label noise. The proposed approach includes node classification module and confident learning module, where the confident learning module selects clean nodes with high confident labels for node classification module to train an accurate model in the graphs with label noise. Pseudo label method is further applied on unlabeled data to increase samples for confident learning and improve the classification performance of subsequent node classification module. We evaluated classification performance of GCN_CL and compared our approach against other models to classify autism spectrum disorders patients versus health controls in population graph, which is constructed from ABIDE I dataset. The experimental results show GCN_CL achieves achieves the best performance in the population graph with different artificial noise levels.

Graph convolutional networks (GCNs) have been successfully applied to many graph data on various learning tasks such as node classification. However, there is limited understanding of the internal logic and decision patterns of GCNs. In this paper, we propose a layer-wise relevance propagation based explanation method for GCNs, namely GCN-LRP, to explore the latent pattern of GCNs. Then, we use three well-known citation network data sets and synthetic graph data sets for node classification tasks with GCN-LRP explanation, and experimentally identify latent attentions when GCNs aggregates information from the node and its neighboring nodes: (i) GCNs pay more attention to the classified node when comparing with its neighbors; (ii) GCNs do not pay attention to all the neighboring nodes equally, and a few neighboring nodes received more attention than others. Moreover, we further theoretically analyze and find that: (i) the latent attentions come from the recursively aggregating of GCNs; (ii) the neighboring nodes, which share enough neighbors with classified node, would receive more attention than other neighbors; (iii) the latent attention could hardly be changed by model training. We also discuss the advantage and limitations of GCNs introduced by the latent attentions, and implications of our findings for graph data learning with GCNs.

BACKGROUND: Advanced sequencing machines dramatically speed up the generation of genomic data, which makes the demand of efficient compression of sequencing data extremely urgent and significant. As the most difficult part of the standard sequencing data format FASTQ, compression of the quality score has become a conundrum in the development of FASTQ compression. Existing lossless compressors of quality scores mainly utilize specific patterns generated by specific sequencer and complex context modeling techniques to solve the problem of low compression ratio. However, the main drawbacks of these compressors are the problem of weak robustness which means unstable or even unavailable results of sequencing files and the problem of slow compression speed. Meanwhile, some compressors attempt to construct a fine-grained index structure to solve the problem of slow random access decompression speed. However, they solve the problem at the sacrifice of compression speed and at the expense of large index files, which makes them inefficient and impractical. Therefore, an efficient lossless compressor of quality scores with strong robustness, high compression ratio, fast compression and random access decompression speed is urgently needed and of great significance. RESULTS: In this paper, based on the idea of maximizing the use of hardware resources, LCQS, a lossless compression tool specialized for quality scores, was proposed. It consists of four sequential processing steps: partitioning, indexing, packing and parallelizing. Experimental results reveal that LCQS outperforms all the other state-of-the-art compressors on all criteria except for the compression speed on the dataset SRR1284073. Furthermore, LCQS presents strong robustness on all the test datasets, with its acceleration ratios of compression speed increasing by up to 29.1x, its file size reducing by up to 28.78%, and its random access decompression speed increasing by up to 2.1x. Additionally, LCQS also exhibits strong scalability. That is, the compression speed increases almost linearly as the size of input dataset increases. CONCLUSION: The ability to handle all different kinds of quality scores and superiority in compression ratio and compression speed make LCQS a high-efficient and advanced lossless quality score compressor, along with its strength of fast random access decompression. Our tool LCQS can be downloaded from https://github.com/SCUT-CCNL/LCQSand freely available for non-commercial usage.

Online mobile advertising plays a vital role in the mobile app ecosystem. The mobile advertising frauds caused by fraudulent clicks or other actions on advertisements are considered one of the most critical issues in mobile advertising systems. To combat the evolving mobile advertising frauds, machine learning methods have been successfully applied to identify advertising frauds in tabular data, distinguishing suspicious advertising fraud operation from normal one. However, such approaches may suffer from labor-intensive feature engineering and robustness of the detection algorithms, since the online advertising big data and complex fraudulent advertising actions generated by malicious codes, botnets, and click-firms are constantly changing. In this paper, we propose a novel weighted heterogeneous graph embedding and deep learning-based fraud detection approach, namely, GFD, to identify fraudulent apps for mobile advertising. In the proposed GFD approach, (i) we construct a weighted heterogeneous graph to represent behavior patterns between users, mobile apps, and mobile ads and design a weighted metapath to vector algorithm to learn node representations (graph-based features) from the graph; (ii) we use a time window based statistical analysis method to extract intrinsic features (attribute-based features) from the tabular sample data; (iii) we propose a hybrid neural network to fuse graph-based features and attribute-based features for classifying the fraudulent apps from normal apps. The GFD approach was applied on a large real-world mobile advertising dataset, and experiment results demonstrate that the approach significantly outperforms well-known learning methods.

BACKGROUND: The advance of next generation sequencing enables higher throughput with lower price, and as the basic of high-throughput sequencing data analysis, variant calling is widely used in disease research, clinical treatment and medicine research. However, current mainstream variant caller tools have a serious problem of computation bottlenecks, resulting in some long tail tasks when performing on large datasets. This prevents high scalability on clusters of multi-node and multi-core, and leads to long runtime and inefficient usage of computing resources. Thus, a high scalable tool which could run in distributed environment will be highly useful to accelerate variant calling on large scale genome data. RESULTS: In this paper, we present ADS-HCSpark, a scalable tool for variant calling based on Apache Spark framework. ADS-HCSpark accelerates the process of variant calling by implementing the parallelization of mainstream GATK HaplotypeCaller algorithm on multi-core and multi-node. Aiming at solving the problem of computation skew in HaplotypeCaller, a parallel strategy of adaptive data segmentation is proposed and a variant calling algorithm based on adaptive data segmentation is implemented, which achieves good scalability on both single-node and multi-node. For the requirement that adjacent data blocks should have overlapped boundaries, Hadoop-BAM library is customized to implement partitioning BAM file into overlapped blocks, further improving the accuracy of variant calling. CONCLUSIONS: ADS-HCSpark is a scalable tool to achieve variant calling based on Apache Spark framework, implementing the parallelization of GATK HaplotypeCaller algorithm. ADS-HCSpark is evaluated on our cluster and in the case of best performance that could be achieved in this experimental platform, ADS-HCSpark is 74% faster than GATK3.8 HaplotypeCaller on single-node experiments, 57% faster than GATK4.0 HaplotypeCallerSpark and 27% faster than SparkGA on multi-node experiments, with better scalability and the accuracy of over 99%. The source code of ADS-HCSpark is publicly available at https://github.com/SCUT-CCNL/ADS-HCSpark.git .

Predicting churning customers in advance allows marketers to retain existing and valuable customers, and to develop a customer churn predicting model is a key issue of customer relationship management in modern marketing. In this paper, a product-based Recurrent Neural Network (pRNN) approach is proposed for customer churn prediction in telecommunication sector. In the proposed model, RNN with long short-term memory units is used to learn sequential patterns from customer data changing over time, and the product operation is introduced before recurrent layer to learn high-order interaction between features. pRNN is applied on a real-world telecommunication dataset; experiment results demonstrate that pRNN significantly outperforms other comparison models.

We propose Top-N-Rank, a novel family of list-wise Learning-to-Rank models for reliably recommending the N top-ranked items. The proposed models optimize a variant of the widely used cumulative discounted gain (DCG) objective function which differs from DCG in two important aspects: (i) It limits the evaluation of DCG only on the top N items in the ranked lists, thereby eliminating the impact of low-ranked items on the learned ranking function; and (ii) it incorporates weights that allow the model to leverage multiple types of implicit feedback with differing levels of reliability or trustworthiness. Because the resulting objective function is non-smooth and hence challenging to optimize, we consider two smooth approximations of the objective function, using the traditional sigmoid function and the rectified linear unit (ReLU). We propose a family of learning-to-rank algorithms (Top-N-Rank) that work with any smooth objective function. Then, a more efficient variant, Top-N-Rank.ReLU, is introduced, which effectively exploits the properties of ReLU function to reduce the computational complexity of Top-N-Rank from quadratic to linear in the average number of items rated by users. The results of our experiments using two widely used benchmarks, namely, the MovieLens data set and the Amazon Video Games data set demonstrate that: (i) The "top-N truncation" of the objective function substantially improves the ranking quality of the top N recommendations; (ii) using the ReLU for smoothing the objective function yields significant improvement in both ranking quality as well as runtime as compared to using the sigmoid; and (iii) Top-N-Rank.ReLU substantially outperforms the well-performing list-wise ranking methods in terms of ranking quality.

Drug-Drug Interaction (DDI) relation extraction is a multi-class classification problem that aims to predict the interaction between drugs in a sentence. The configuration of Convolutional Neural Network (CNN) in relation extraction usually applied shallow architecture layers, which may make the information in given input text is not fully captured, thus fail to capture a long sentence containing the detected drug relation or some irrelevant word captured during the feature extraction process. This paper proposed an extending depth of the CNN layer called DeepCNN for DDI relation extraction. The DeepCNN learns the high quality of the learning representation so that it is able to cover long input sentences as the typical of DDIExtraction dataset. We use multi-channel word-embedding to enlarge the vocabulary and decrease the number of unknown words, and Adam update rule to automatically learn the network parameters of DeepCNN for DDI relation extraction. The experiments show that the architecture of 10 layers DeepCNN successfully obtained the significant improvement compared to the previous CNN method in DDI relation extraction. The result proves that CNN is a robust and well-deserved for DDI relation extraction.

Due to the randomness and noisiness, quality scores presented in sequencing data have already comprised about 70% of the compressed storage and reached their lossless compression limit. Lossy compression of quality scores, which guarantees the performance of subsequent variant calling procedure, has been an ideal candidate and a great challenge in big genomic data analysis. Currently, state-of-the-art locality-based lossy compressor PBlock, based on the assumption that all the quality score lines should exhibit a single locality, applies identical and static locality criterion manually to smooth all the different quality score lines. However, this assumption is usually not the real case and would inevitably result in sub-optimal locality criteria in some quality score lines, which eventually leads to performance degradation of lossy compression and variant calling procedure. Therefore, on the basis of a more reasonable assumption that different quality score lines should exhibit different locality, an enhanced version of lossy compressor PBlock called ALL-CQS is proposed. In this paper, ALL-CQS applies adaptive locality criteria to smooth different quality ality scores lines automatically based on PBlock' lossy mechanism. Experimental results reveal that our lossy compressor ALL-CQS not only achieves the best variant calling performance which is very close to the lossless one, but also outperforms all the other state-of-the-art lossy compressors and achieves up to 145% improvements over the original lossless compressors in terms of compression ratio.

With the rapid development of the mobile games, mobile game recommendation has become a core technique to mobile game marketplaces. This paper proposes a bipartite graph based recommendation algorithm PKBBR (Prior Knowledge Based for Bipartite Graph Rank). We model the user's interest in mobile game based on bipartite graph structure and use the users' mobile game behavior to describe the edge weights of the graph, then incorporate users' prior knowledge into the projection procedure of the bipartite graph to enrich the information among the nodes. Because the popular games have a great influence on mobile game marketplace, we design a hybrid recommendation algorithm to incorporate popularity recommendation based on users' behaviors. The experiment results show that this hybrid method could achieve a better performance than other approaches.

In this paper, we propose a lossless Quality Scores compression methods of FASTQ file format which is commonly used to store the NGS (Next Generation Sequencing). Instead of elaborating excellent data structure and compression technique based on the original FASTQ Quality Scores file, we try to change the distribution of original FASTQ Quality Scores file through linear combination prediction so as to make it better for further compression, using existing compression algorithms. Experimental results indicate that our compression method outperforms other four state-of-the-art compression tools and achieves up to 10% improvement in compression ratio comparing with the one in the second place and up to 35% over Gzip both on on all test datasets.

In this paper, we propose a non-reference based and lossless compression tool of FASTQ which is commonly used to store the NGS. Instead of elaborating excellent data structure and compression technique based on the original FASTQ file, we try to change the distribution of original FASTQ file so as to make it better for further compression by existing compression tools. Experimental results indicate that our method outperforms all the six state-of-the-art compression tools and achieves up to 10% ∼ 43% improvement in terms of the average compression ratio. Besides, our compression tool LCTD outperforms Fastqz in both compression ratio and speed and the latter compression tool Fastqz wins the world champion of compression competition SequenceSqueeze. The source program is available by sending email to us.

The uncertainty and extreme large scale of cloud resources make task scheduling very difficult which affects the user quality of experience and probably result in a waste of cloud resources and energy consumption. Moreover, some resources stay in an unusable state for extended time. To take into account these problems a cloud resource evaluation model is proposed, termed Entropy Optimization Evaluation and ant colony clustering Model (EOEACCM). The model releases long-term unavailable resources to save energy. First, by mean of the entropy increasing minimum principle, the proposed model can maximize the system utilization and balance profits of both cloud resource providers and users. As a consequence, it can shorten task completion time. Secondly, the model narrows the task scheduling size and achieves optimal scheduling by clustering. To make the model more suitable for the dynamics of cloud resources, the model design improves pheromone update policies by fixing total path length in each function cycle when clustering by the ant colony algorithm. Evaluation of results using EOEACCM demonstrate that it may be applicable for resource management strategies for migration and release, an application which can effectively save energy. The proposed model was evaluated by simulation. Experiment results showed the positive effect of user satisfaction from entropy optimization, as well as scheduling time from clustering. Moreover, when the scale of tasks was large, this clustering algorithm performed much better than others. The clustering model also demonstrated better adaptability when some cloud resources were joined or terminated.

With the large-scale deployment of virtualized data centers, energy consumption and SLA (Service Level Agreement) violation have already become the urgent issue to be solved. And it is essential and important to design energy-aware allocation policy for energy-aware and SLA violation reduction. In this paper, we propose a novel allocation and selection policy for the dynamic virtual machine (VM) consolidation in virtualized data centers to reduce energy consumption and SLA violation. Firstly, we use the mean and standard deviation of CPU utilization for VM to determine the hosts overloaded or not, secondly we use the positive maximum correlation coefficient to select VMs from those overloading hosts for migration. Although the proposed allocation and selection policies performs a little worse than the previous ones in energy consumption, experiments show that it performs greatly better than the previous ones on the whole.

Grid system has many great security challenges such as access control. The attribute-based access control model (ABAC) has much merits that are more flexible, fine-grained and dynamically suitable to grid environment. As an important factor in grid security, trust is increasingly applied to management of security, especially in access control. This paper puts forward a novel trust model in multi-domain grid environment and trust factor was originally introduced into access control architecture of grid to extend classic ABAC model. By extending the authorization architecture of XACML, extended ABAC based access control architecture for grid was submitted. In our experiment, the increase and decrease of trust are non-symmetrical and the trust model is sensitive to the malicious attacks. It can effectively control the trust change of different nodes and the trust model can reduce effectively the damage of vicious attack.

The cellular Potts model (CPM) has been widely used for biological simulations. However, most of current implementations are either sequential or approximate, which cannot be used for large scale complex 3D simulation. In this paper we present a hybrid parallel framework for CPM simulations. The time-consuming partial differential equation (PDE) solving, cell division, and cell reaction operation are distributed to clusters by using the message passing interface (MPI). The Monte Carlo lattice update is parallelized on shared-memory SMP system by using OpenMP. Since the Monte Carlo lattice update is much faster than the PDE solving and SMP systems are more and more common, this hybrid approach achieves good performance and high accuracy at the same time. Based on the parallel cellular Potts model, we have studied the avascular tumor growth by using a multiscale model. The application and performance analyses demonstrate that the hybrid parallel framework is quite efficient. The hybrid parallel CPM can be used for the large scale simulation (~ 108sites) of complex collective behavior of numerous cells (~ 106).

Receptor-based virtual screening becomes more and more important in novel leads design in pharmaceutical industry. However, virtual screening technology is not mature enough and still growing with the dynamic development of related technology and science. So a flexible way to organize multiple biological algorithms and distributed chemical databases to fulfill virtual screening process is in great expect. Base on service-oriented architecture (SOA) and grid technology, we have developed a service oriented virtual screening platform that enables dynamic algorithms integrating. On which biological user can go from amino acid sequence to candidate drug hits without paying too much attention to the low-level operation, also the platform is easy to extend when new algorithms come. The receptor predicting accuracy is enhanced while effectively combining homology modeling and Ab initio method, this lead to virtual screening improvement. Through the performance test experiments, it is observed that SOA approach obtained a somewhat satisfying performance and demonstrates that SOA pattern's advantage highly surpass the performance overhead.

Finite element method (FEM)-based scientific numerical simulations are often computing-extensive and grid platform can speed up the simulation progress significantly. However, the coupling of FEM-based simulation tasks with grid platform isn't so straightforward. Adopting the service-oriented architecture (SOA), we develop a general service-oriented grid computing platform for FEM-based simulations (GSGCP-FEM). GSGCP-FEM provides users with the ability to make a general FEM-based simulation to be service-oriented. Basing on the services provided by GSGCP-FEM and grid middleware, users can deploy new scientific simulation applications easily. In this paper, we explain the design, architecture, and implementation of GSGCP-FEM in detail. We also deploy two practical applications based on GSGCP-FEM to demonstrate its usefulness

With the explosion of low-cost bandwidth availability, many desirable e-business features require new functionality in enterprise networking systems, and thus force the network components to be both flexible and powerful. Programmable routers (typically NP-based systems) as a platform on which to implement the enterprise applications seem to fit the bill. Meanwhile, the demands for accommodating an increasingly diverse range of novel network services in such system are still keep growing. To meet these challenges, this paper presents a novel methodology for building control plane software using component technology in programmable network environment. The proposed approach can facilitate the developing of scalable, flexible modular NP-based systems that are required to the e-business world

Ad hoc networks are a new kind of mobile computer network with broad applications and important commercial values. Most proposed on-demand routing protocols for Ad hoc networks build and rely on single path. However, the single path is easily broken and needs to perform a route discovery process again due to the dynamic topology of ad hoc networks. In Aa hoc networks multipath routing is better suited than single path in stability and load balance. Our interests lie in how to get a stable route that can be used for a longer time without rerouting to recover from the path breakage. Our analysis of the multipath's stability shows that the stability of multipath routing is closely related to the routing policies. An independent stability-based routing scheme that is based on DSR is then presented together with the utilization of a group of independent paths for routing. The stable multipath embedded in our muting policies can enhance the performance by decreasing rerouting overheads. The simulation results show that the proposed algorithm can be more adaptive the mobile environment and outperform SMR and DSR.