Predicting drug–drug interactions (DDIs) is crucial for understanding and preventing adverse drug reactions (ADRs). However, most existing methods inadequately explore the interactive information between drugs in a self-supervised manner, limiting our comprehension of drug–drug associations. This paper introduces EDDINet : E nhancing D rug- D rug I nteraction Prediction via Information Flow and Consensus-Constrained Multi-Graph Contrastive Learning for precise DDI prediction. We first present a cross-modal information-flow mechanism to integrate diverse drug features, enriching the structural insights conveyed by the drug feature vector. Next, we employ contrastive learning to filter various biological networks, enhancing the model’s robustness. Additionally, we propose a consensus regularization framework that collaboratively trains multi-view models, producing high-quality drug representations. To unify drug representations derived from different biological information, we utilize an attention mechanism for DDI prediction. Extensive experiments demonstrate that EDDINet surpasses state-of-the-art unsupervised models and outperforms some supervised baseline models in DDI prediction tasks. Our approach shows significant advantages and holds promising potential for advancing DDI research and improving drug safety assessments. Our codes are available at: https://github.com/95LY/EDDINet_code . • EDDINet is proposed via Information Flow and Consensus-Constrained Multi-Graph Contrastive Learning. • Introduce the cross-modal information-flow mechanism to facilitate the information across various drug features. • EDDINet surpasses state-of-the-art unsupervised models and even outperforms some supervised baseline models in DDI prediction tasks.

MicroRNAs (miRNAs) are critical in diagnosing and treating various diseases. Automatically demystifying the interdependent relationships between miRNAs and diseases has recently made remarkable progress, but their fine-grained interactive relationships still need to be explored. We propose a multi-relational graph encoder network for fine-grained prediction of miRNA-disease associations (MRFGMDA), which uses practical and current datasets to construct a multi-relational graph encoder network to predict disease-related miRNAs and their specific relationship types (upregulation, downregulation, or dysregulation). We evaluated MRFGMDA and found that it accurately predicted miRNA-disease associations, which could have far-reaching implications for clinical medical analysis, early diagnosis, prevention, and treatment. Case analyses, Kaplan-Meier survival analysis, expression difference analysis, and immune infiltration analysis further demonstrated the effectiveness and feasibility of MRFGMDA in uncovering potential disease-related miRNAs. Overall, our work represents a significant step toward improving the prediction of miRNA-disease associations using a fine-grained approach could lead to more accurate diagnosis and treatment of diseases.

Early detection significantly enhances patients' survival rates by identifying tumors in their initial stages through medical imaging. However, prevailing methodologies encounter challenges in extracting comprehensive information from diverse modalities, thereby exacerbating semantic disparities and overlooking critical task correlations, consequently compromising the accuracy of prognosis predictions. Moreover, clinical insights emphasize the advantageous sharing of parameters between tumor segmentation and survival prediction for enhanced prognostic accuracy. This paper proposes a novel model, BTSSPro, designed to concurrently address Breast cancer Tumor Segmentation and Survival prediction through a Prompt-guided multi-modal co-learning framework. Technologically, our approach involves the extraction of tumor-specific discriminative features utilizing shared dual attention (SDA) blocks, which amalgamate spatial and channel information from breast MR images. Subsequently, we employ a guided fusion module (GFM) to seamlessly integrate the Electronic Health Record (EHR) vector into the extracted tumor-related discriminative feature representations. This integration prompts the model's feature selection to align more closely with real-world scenarios. Finally, a feature harmonic unit (FHU) is introduced to coordinate the transformer encoder and CNN decoder, thus reducing semantic differences. Remarkably, BTSSPro achieved a C-index of 0.968 and Dice score of 0.715 on the Breast MRI-NACT-Pilot dataset and a C-index of 0.807 and Dice score of 0.791 on the ISPY1 dataset, surpassing the previous state-of-the-art methods.

Cancer heterogeneity has posed great challenges in exploring precise therapeutic strategies for cancer treatment. The identification of cancer subtypes aims to detect patients with distinct molecular profiles and thus could provide new clues on effective clinical therapies. While great efforts have been made, it remains challenging to develop powerful computational methods that can efficiently integrate multi-omics datasets for the task. In this paper, we propose a novel self-supervised learning model called Deep Multi-view Contrastive Learning (DMCL) for cancer subtype identification. Specifically, by incorporating the reconstruction loss, contrastive loss and clustering loss into a unified framework, our model simultaneously encodes the sample discriminative information into the extracted feature representations and well preserves the sample cluster structures in the embedded space. Moreover, DMCL is an end-to-end framework where the cancer subtypes could be directly obtained from the model outputs. We compare DMCL with eight alternatives ranging from classic cancer subtype identification methods to recently developed state-of-the-art systems on 10 widely used cancer multi-omics datasets as well as an integrated dataset, and the experimental results validate the superior performance of our method. We further conduct a case study on liver cancer and the analysis results indicate that different subtypes might have different responses to the selected chemotherapeutic drugs.

MOTIVATION: Molecular property prediction is a significant requirement in AI-driven drug design and discovery, aiming to predict the molecular property information (e.g. toxicity) based on the mined biomolecular knowledge. Although graph neural networks have been proven powerful in predicting molecular property, unbalanced labeled data and poor generalization capability for new-synthesized molecules are always key issues that hinder further improvement of molecular encoding performance. RESULTS: We propose a novel self-supervised representation learning scheme based on a Cascaded Attention Network and Graph Contrastive Learning (CasANGCL). We design a new graph network variant, designated as cascaded attention network, to encode local-global molecular representations. We construct a two-stage contrast predictor framework to tackle the label imbalance problem of training molecular samples, which is an integrated end-to-end learning scheme. Moreover, we utilize the information-flow scheme for training our network, which explicitly captures the edge information in the node/graph representations and obtains more fine-grained knowledge. Our model achieves an 81.9% ROC-AUC average performance on 661 tasks from seven challenging benchmarks, showing better portability and generalizations. Further visualization studies indicate our model's better representation capacity and provide interpretability.

Automated detection of depression using Electroencephalogram (EEG) signals has become a promising application in advanced bioinformatics technology. Although current methods have achieved high detection performance, several challenges still need to be addressed: (1) Previous studies do not consider data redundancy when modeling multi-channel EEG signals, resulting in some unrecognized noise channels remaining. (2) Most works focus on the functional connection of EEG signals, ignoring their spatial proximity. The spatial topological structure of EEG signals has not been fully utilized to capture more fine-grained features. (3) Prior depression detection models fail to provide interpretability. To address these challenges, this paper proposes a new model, Multi-view Graph Contrastive Learning via Adaptive Channel Optimization (MGCL-ACO) for depression detection in EEG signals. Specifically, the proposed model first selects the critical channels by maximizing the mutual information between tracks and labels of EEG signals to eliminate data redundancy. Then, the MGCL-ACO model builds two similarity metric views based on functional connectivity and spatial proximity. MGCL-ACO constructs the feature extraction module by graph convolutions and contrastive learning to capture more fine-grained features of different perspectives. Finally, our model provides interpretability by visualizing a brain map related to the significance scores of the selected channels. Extensive experiments have been performed on public datasets, and the results show that our proposed model outperforms the most advanced baselines. Our proposed model not only provides a promising approach for automated depression detection using optimal EEG signals but also has the potential to improve the accuracy and interpretability of depression diagnosis in clinical practice.

Predicting microbe-disease associations is crucial for demystifying the causes of diseases and preventing them proactively. However, most of existing approaches are feeble to comprehensively investigate the interactive relationships between diseases and microbes by self-supervised manner, resulting in the microbe-disease associations are hard to ploughed. In this paper, we propose DNCL-MDA, a novelMicrobe-DiseaseAssociations prediction model based onDualNetworkContrastiveLearning, to demystify potential microbe-disease associations (MDAs). Particularly, DNCL-MDA first constructs a pair of microbe-disease dual networks, and designs an effective information flow projection method to obtain the individual disease and microbe networks while reserving their interdependent information. Then, DNCL-MDA proposes an optimized graph contrastive learning approach to learn the discriminative feature representations of diseases and microbes. Finally, the feature representations are contacted and fed into a fully connected neural network to predict the potential microbe-disease associations automatically. Experimental results on real-world datasets demonstrate that our proposed DNCL-MDA largely outperforms the state-of-the-art methods with more promising AUC performances.

Cancer subtype identification is to classify cancer into groups according to their molecular characteristics and clinical manifestations and is the basis for more personalized diagnosis and therapy. Public datasets such as The Cancer Genome Atlas (TCGA) have collected a massive number of multi-omics data. The accumulation of these datasets provides unprecedented opportunities to study the mechanism of cancers and further identify cancer subtypes at a comprehensive level. In this paper, we propose a multi-view robust graph-based clustering (MRGC) method to effectively identify cancer subtypes. Our method first learns robust latent representations from the raw omics data to alleviate the influences of the noise, where a set of similarity matrices are then adaptively learned based on these new representations. Finally, a global similarity graph is obtained by exploiting the consensus structure from the graphs. As a result, the three parts in our method can reinforce each other in a mutual iterative manner. We conduct extensive experiments on both generic machine learning datasets and cancer datasets. The experimental results confirm that our model can achieve satisfactory clustering performance compared to several state-of-the-art approaches. Moreover, we convey the practicability of MRGC by carrying out a case study on hepatocellular carcinoma.

Increasing biological evidence indicated that microRNAs (miRNAs) play a vital role in exploring the pathogenesis of various human diseases (especially in tumors). Mining disease-related miRNAs is of great significance for the clinical diagnosis and treatment of diseases. Compared with the traditional experimental methods with the significant limitations of high cost, long cycle and small scale, the methods based on computing have the advantages of being cost-effective. However, although the current methods based on computational biology can accurately predict the correlation between miRNAs and disease, they can not predict the detailed association information at a fine level. We propose a knowledge-driven approach to the fine-grained prediction of disease-related miRNAs (KDFGMDA). Different from the previous methods, this method can finely predict the clear associations between miRNA and disease, such as upregulation, downregulation or dysregulation. Specifically, KDFGMDA extracts triple information from massive experimental data and existing datasets to construct a knowledge graph and then trains a depth graph representation learning model based on knowledge graph to complete fine-grained prediction tasks. Experimental results show that KDFGMDA can predict the relationship between miRNA and disease accurately, which is of far-reaching significance for medical clinical research and early diagnosis, prevention and treatment of diseases. Additionally, the results of case studies on three types of cancers, Kaplan-Meier survival analysis and expression difference analysis further provide the effectiveness and feasibility of KDFGMDA to detect potential candidate miRNAs. Availability: Our work can be downloaded from https://github.com/ShengPengYu/KDFGMDA.

Abstract Large collections of electronic clinical data today provide us with a vast source of information on medical practice. However, the utilization of those data for exploratory analysis to support clinical decisions is still limited. It is particularly challenging for extracting useful disease progression patterns from such data because it is longitudinal, incomplete, irregular, and heterogeneous of the patient conditions. In this article, we propose an integrated clinical event prediction model medical concept integrated residual short‐long temporal convolutional networks (SL‐TCN) to address these challenges. Compared to existing models, our model has three‐fold advantages: (1) it learns a compact set of medical concepts as the bridge between the hidden progression process and the observed medical evidence. (2) it learns a continuous‐time progression model from discrete‐time observations with nonequal intervals and high‐dimensional features. (3) We fuse the temporal convolutional network, the long short‐term memory network, and the residual connector so as to capture the local and global dependency of the sequence and make the clinical event predictions more robust. Through extensive experiments on the MIMIC III dataset, we demonstrate that our SL‐TCN achieves higher precision in clinical event prediction and derives some interesting clinical insights.

Summary The evaluation of Internet advertising effectiveness plays an important role in network marketing at this stage. The evaluation criteria of Internet advertising effectiveness mostly use proceeds as an indicator, ignoring the brand effect generated by Internet advertising, making it difficult for researchers to further study the effect of Internet advertising. With the updating and iteration of technical theory, it is an urgent task to find a new criteria to measure Internet advertising effectiveness. At present, the research on Internet advertising effectiveness research is based on the analysis of nondifferentiation category. The research on user click behavior is a holistic analysis of the whole web page. This global coverage research is difficult to present the behavior characteristics of users in the web page. In response to the above problems, the impression space is proposed to measure Internet advertising effectiveness. Firstly, the fuzzy classification theory is used to digitize the user category data, the user type is divided by probability distribution, and the distribution of web page interest regions is divided by gaze time and page layout to further study the click‐and‐effect relationship of different users and the user browsing time series.Then, establish multimodal multifactors to build an impression space model to measure Internet advertising effectiveness. The experimental results show that the accuracy of the constructed impression space model for the evaluation of Internet advertising effectiveness reaches 92.4%, and the impression space is more effective as the evaluation of Internet advertising effectiveness.

Web ad effect evaluation is a challenging problem in web marketing research. Although the analysis of web ad effectiveness has achieved excellent results, there are still some deficiencies. First, there is a lack of an in‐depth study of the relevance between advertisements and web content. Second, there is not a thorough analysis of the impacts of users and advertising features on user browsing behaviors. And last, the evaluation index of the web advertisement effect is not adequate. Given the above problems, we conducted our work by studying the observer's behavioral pattern based on multimodal features. First, we analyze the correlation between ads and links with different searching results and further assess the influence of relevance on the observer's attention to web ads using eye‐movement features. Then we investigate the user's behavioral sequence and propose the directional frequent‐browsing pattern algorithm for mining the user's most commonly used browsing patterns. Finally, we offer the novel use of “memory” as a new measure of advertising effectiveness and further build an advertising memory model with integrated multimodal features for predicting the efficacy of web ads. A large number of experiments have proved the superiority of our method.