Cell-cell communication plays a pivotal role in tissue development, homeostasis, and disease progression. However, existing approaches often lack systematic modeling of networklevel structure and often overlook spatial modular organization. To address these limitations, we propose CCNVGAE, an unsupervised framework based on variational graph autoencoder (VGAE) that integrates spatial coordinates, ligand-receptor gene expression and graph structural information for functional module identification. Experiments on spatial transcriptomic datasets from mouse brain and human lung cancer demonstrate that CCNVGAE achieves competitive performance in spatial segmentation and biological interpretability. Moreover, it effectively uncovers module specific ligand-receptor interactions and their potential regulatory mechanisms, providing a powerful tool for elucidating spatial communication structures in complex tissues.

Accurately annotating non-coding regulatory DNA, critical for gene regulation and implicated in numerous diseases, remains a major challenge. Current deep learning models struggle to integrate local motifs with long-range dependencies due to limited architectures and simplistic feature fusion. We propose DCAM, a dual-branch neural network where a global branch captures long-range context via a CNN-BiLSTM, while a local branch extracts fine-grained motifs using dna2vec and CNNs. These features are dynamically integrated via a sequential channel-spatial attention mechanism. Evaluated on diverse benchmark datasets, DCAM significantly outperforms state-of-the-art models. DCAM also accurately predicts the functional impact of single nucleotide variants (SNVs), establishing it as a powerful tool for prioritizing pathogenic non-coding variants.

The polygenic, nonlinear, and dynamical regulation of complex traits presents a significant challenge in accurately identifying cell states linked to these traits. Traditional methods that average expression at the cell-type level fail to capture singlecell heterogeneity, while those focusing solely on gene expression overlook the impact of intricate gene regulatory networks. This study introduces NetCell, a network-based approach for pinpointing cell populations associated with traits at single-cell resolution. By leveraging gene interaction networks, NetCell enhances the precision and interpretability of trait-associated cell identification and facilitates the discovery of pivotal genes. NetCell successfully identified crucial asthma-associated genes by pinpointing relevant tissue and cell types.

Complex diseases are determined by many genetic and environmental factors. The Omnigenic model provides a new conceptual model for exploring the pathogenic mechanism of complex diseases. The model points out that the weak genetic effects of a large number of peripheral genes are superimposed on the core genes, thus contributing most of the heritability to complex diseases. At present, many studies have explored and verified the important role of peripheral genes. However, few studies have focused on the impact of peripheral genes on the regulation of core genes in complex diseases. Therefore, we proposed an embedding tool called PToC to characterize the impact of the weak genetic effects of peripheral genes on the core genes of the disease, and represent the impact into weak effect vectors from the two aspects of network topological structure and biological functional. In the case study on schizophrenia, we curated core and peripheral genes. First, we extracted weak effect vectors using PToC, and applied them to identify peripheral genes, achieved an average AUC 17.90% higher than other network embedding methods. Additionally, we constructed a sample classifier based on weak effect vectors. Experimental results show that the average AUC value of this method in distinguishing different sample types is 0.91, which is better than existing methods. Thus, PToC offers a novel perspective and tool for analyzing the weak effects of peripheral genes, enhancing disease diagnosis and treatment.

The rise of single-cell RNA sequencing (scRNA-seq) technology allows us to capture transcriptomic data at the individual cell level. Utilizing unsupervised clustering based on transcriptome similarity aids in accurately identifying essential cell types. However, challenges such as high dimensionality, noise, and data sparsity within scRNA-seq pose difficulties for conventional similarity assessment. Addressing these challenges is crucial for understanding heterogeneity among cell populations. Consequently, we propose a novel approach, CellSim, for assessing cell-cell similarity based on network propagation. We evaluated the method’s performance by clustering cells across 11 scRNA-seq datasets. Simultaneously, we conducted simulation experiments to assess the method’s ability to identify relationships between cells in datasets with high dropout rates. The results indicated that CellSim is superior to others in terms of evaluating clustering results. Eventually, we apply these similarity methods to the same trajectory inference framework on embryo datasets. The results reveal that our method outperforms others in inferring the cell trajectory and exhibits strong consistency with the ground truth.

Systematically, network pharmacology based on homogeneous interaction network of encoding genes has effectively elucidated treatment effects, multiple actions, and associated side effects of drugs by quantifying the proximity between disease modules and drug modules. However, further exploration is required to understand the contributions of heterogeneous and multi-scale factors, including non-coding genes and GO Terms, to proximity. This study aims to integrate a heterogeneous interaction network(HIN) consisting of encoding genes, non-coding genes, and GO Terms. We find that incorporating non-coding genes and GO Terms significantly enhances the proximity between drug modules and the diseases modules they address. Additionally, we propose an index for quantifying crucial nodes based on network proximity, enabling the identification of treatment key coding genes, non-coding genes, and GO terms (Tkey) involved in drug effect processes. In a systematic validation of 712 drug-disease pairs encompassing 471 drugs and 35 diseases, we effectively identify heterogeneous and multi-scale Tkeyfactors that play crucial roles in elucidating the mechanisms underlying drug treatment for various diseases. As a case study, we present Tkeyand elucidate the pivotal biological pathways involved in Theophylline treatment for Asthma, multiple Theophylline actions of Theophylline for Pulmouary Emphysema, and the pathways leading to the side effect of Hypotension induced by Theophylline. In conclusion, heterogeneous and multi-scale Tkeyare imperative for mediating drug effects in interaction networks, thus treating diseases or causing side effects. Our research presents a feasible computational tool for understanding how drugs treat diseases in drug-disease relationship analysis.

Cancer development is associated with thousands of coding and non-coding genes. Understanding the role of non-coding RNAs (ncRNAs) in cancer neighbourhood is of great significance for interpreting the occurrence and diagnosis of cancer. This paper first constructs a heterogeneous interaction network containing coding and non-coding genes. Then, the performance of network connectivities when applied to cancer-perturbed genes is evaluated by statistical index. We found that ncRNAs considerably improve the connectivity of perturbed genes and mediate more genes into cancer neighbourhood. Additionally, a method connected Perturbed Region (CPR) for mining cancer neighbourhood with the participation of ncRNAs in the network is proposed based on the bridging role of ncRNAs among cancer perturbed genes. Furthermore, we observed the role of ncRNAs in biologically complementing and promoting the formation of cancer-related pathways. Our results topologically promote the pattern of ncRNAs in forming cancer and biologically are of great benefit to reveal a more accurate disease relationship.

Two thousand nineteen novel coronavirus SARS-CoV-2, the pathogen of COVID-19, has caused a catastrophic pandemic, which has a profound and widespread impact on human lives and social economy globally. However, the molecular perturbations induced by the SARS-CoV-2 infection remain unknown. In this paper, from the perspective of omnigenic, we analyze the properties of the neighborhood perturbed by SARS-CoV-2 in the human interactome and disclose the peripheral and core regions of virus-host network (VHN). We find that the virus-host proteins (VHPs) form a significantly connected VHN, among which highly perturbed proteins aggregate into an observable core region. The non-core region of VHN forms a large scale but relatively low perturbed periphery. We further validate that the periphery is non-negligible and conducive to identifying comorbidities and detecting drug repurposing candidates for COVID-19. We particularly put forward a flower model for COVID-19, SARS and H1N1 based on their peripheral regions, and the flower model shows more correlations between COVID-19 and other two similar diseases in common functional pathways and candidate drugs. Overall, our periphery-core pattern can not only offer insights into interconnectivity of SARS-CoV-2 VHPs but also facilitate the research on therapeutic drugs.

Single-cell clustering is an important part of analyzing single-cell RNA-sequencing data. However, the accuracy and robustness of existing methods are disturbed by noise. One promising approach for addressing this challenge is integrating pathway information, which can alleviate noise and improve performance. In this work, we studied the impact on accuracy and robustness of existing single-cell clustering methods by integrating pathways. We collected 10 state-of-the-art single-cell clustering methods, 26 scRNA-seq datasets and four pathway databases, combined the AUCell method and the similarity network fusion to integrate pathway data and scRNA-seq data, and introduced three accuracy indicators, three noise generation strategies and robustness indicators. Experiments on this framework showed that integrating pathways can significantly improve the accuracy and robustness of most single-cell clustering methods.

BACKGROUND: Precise disease module is conducive to understanding the molecular mechanism of disease causation and identifying drug targets. However, due to the fragmentization of disease module in incomplete human interactome, how to determine connectivity pattern and detect a complete neighbourhood of disease based on this is still an open question. RESULTS: In this paper, we perform exploratory analysis leading to an important observation that through a few intermediate nodes, most separate connected components formed by disease-associated proteins can be effectively connected and eventually form a complete disease module. And based on the topological properties of these intermediate nodes, we propose a connect separate connected components (C3) method to detect a succinct disease module by introducing a relatively small number of intermediate nodes, which allows us to obtain more pure disease module than other methods. Then we apply C3 across a large corpus of diseases to validate this connectivity pattern of disease module. Furthermore, the connectivity of the perturbed genes in multi-omics data such as The Cancer Genome Atlas also fits this pattern. CONCLUSIONS: C3 tool is not only useful in detecting a clearly-defined connected disease neighbourhood of 299 diseases and cancer with multi-omics data, but also helpful in better understanding the interconnection of phenotypically related genes in different omics data and studying complex pathological processes.

Identifying driver modules or pathways is a key challenge to interpret the molecular mechanisms and pathogenesis underlying cancer. An increasing number of studies suggest that rarely mutated genes are important for the development of cancer. However, the driver modules consisting of mutated genes with low-frequency driver mutations are not well characterized. To identify driver modules with rarely mutated genes, we propose a functional similarity index to quantify the functional relationship between rarely mutated genes and other ones in the same module. Then, we develop a method to detect Driver Modules with Rarely mutated Genes (DMRG) by incorporating the functional similarity, coverage and mutual exclusivity. By applying DMRG on TCGA cancer dataset on three networks: HINT+HI2012, iRefIndex and MultiNet, we detect driver modules intersecting with the well-known signalling pathways and protein complexes, such as the cell cycle pathway and the mediator complex. DMRG can also detect driver modules effectively with 20, 40, 60 and 80 percent of samples by random selection. When compared with HotNet2, DMRG detects more rarely mutated cancer genes and has higher pathway enrichment. Overall, DMRG provides an effective method for the identification of driver modules with rarely mutated genes.

Computational approaches for predicting drug-disease associations by integrating gene expression and biological network provide great insights to the complex relationships among drugs, targets, disease genes, and diseases at a system level. Hepatocellular carcinoma (HCC) is one of the most common malignant tumors with a high rate of morbidity and mortality. We provide an integrative framework to predict novel d rugs for HCC based on multi-source random walk (PD-MRW). Firstly, based on gene expression and protein interaction network, we construct a gene-gene weighted i nteraction network (GWIN). Then, based on multi-source random walk in GWIN, we build a drug-drug similarity network. Finally, based on the known drugs for HCC, we score all drugs in the drug-drug similarity network. The robustness of our predictions, their overlap with those reported in Comparative Toxicogenomics Database (CTD) and literatures, and their enriched KEGG pathway demonstrate our approach can effectively identify new drug indications. Specifically, regorafenib (Rank = 9 in top-20 list) is proven to be effective in Phase I and II clinical trials of HCC, and the Phase III trial is ongoing. And, it has 11 overlapping pathways with HCC with lower p-values. Focusing on a particular disease, we believe our approach is more accurate and possesses better scalability.

Three‐dimensional (3D) facial data offer the potential to overcome the difficulties caused by the variation of head pose and illumination in 2D face recognition. In 3D face recognition, localisation of nose tip is essential to face normalisation, face registration and pose correction etc. Most of the existing methods of nose tip detection on 3D face deal mainly with frontal or near‐frontal poses or are rotation sensitive. Many of them are training‐based or model‐based. In this study, a novel method of nose tip detection is proposed. Using pose‐invariant differential surface features – high‐order and low‐order curvatures, it can detect nose tip on 3D faces under various poses automatically and accurately. Moreover, it does not require training and does not depend on any particular model. Experimental results on GavabDB verify the robustness and accuracy of the proposed method.

We present a heuristic global network alignment algorithm, called Alignment based Multiple Hubs(AMH). AMH is efficient in resolving the fatal problem of most conventional algorithms that the initialization selected seeds have a direct influence on the alignment result. This algorithm outperforms state-of-the-art algorithms at detecting conserved functional modules and retrieves in particular 86% more conserved interactions than IsoRank.