Bioinformatics and Genomic Networks

AutoGenome V2 integrates multi-omics data using a novel multimodal deep learning approach, achieving state-of-the-art results in four tasks.

The paper detects putative disease modules by combining data-driven and hypothesis-driven approaches using significant shortest paths and biological pathways in gene expression data from inflammatory bowel disease.

CeTF identifies transcription factors regulating gene networks using PCIT and RIF analysis on gene expression data from various platforms.

The paper constructs a tissue- and organ-based subcellular-function atlas of obesity-related genes and pathways using publicly available datasets and integration methods.

The paper builds a structure-supported human interactome by integrating structural information with PPI data, identifying domain-domain and peptide-binding interactions to improve the resolution and biological relevance of the network.

The paper introduces a network-based method to predict cancer driver genes by analyzing gene importance in network anomaly propagation, showing improved accuracy and precision over existing methods.

This paper introduces NESS, a tool that improves gene-disease association discovery by leveraging multi-species functional genomics data and network analysis to overcome data sparsity and bias.

This paper introduces an algorithm that identifies co-expression of bio-strings from protein sequences and validates results using existing databases.

The paper identifies brain-wide and region-specific gene co-expression modules using RNA-seq data and links them to disease risk variants and biological processes.

The paper introduces MAGINE, a framework for analyzing multi-omics time-course data to build mechanistic hypotheses of cellular responses to perturbations.

The paper introduces a new statistical method, BRT, to assess biological relevance in omics data by incorporating effect size with p-values, and demonstrates its application on microarray data.

The paper introduces a new network mining algorithm to identify pathway-centric interactions in cancer data, revealing dense between-pathway interactions with therapeutic potential.

The paper identifies a mean-correlation relationship in RNA-seq data that biases co-expression analysis and introduces SpQN to correct this bias in network reconstruction.

This paper uses graph embedding methods to prioritize candidate genes in rice by integrating gene data from multiple databases and evaluating link prediction performance.

The paper introduces Maestro, a network analysis method that identifies synchronized genetic activities in Alzheimer's disease, finding a pattern associated with AD that is significantly more common in cases than controls.

The paper uses an integrated data mining approach to identify boolean relationships between Src and its targets by analyzing genomic, literature, and signaling databases.

The paper evaluates how network and gene expression data affect ProTINA's accuracy, finding that network data dominates and introduces TREAP, a new algorithm with higher accuracy.

The paper shows that removing low-degree nodes from a gene interaction network improves prediction of expression dependencies compared to a random model.

The paper develops a method to analyze gene expression and connectivity in Alzheimer's disease hippocampus data, finding heterogeneous gene changes but consistent pathway patterns across datasets.

This paper reanalyzes T2D GWAS data using pathway and network approaches to identify molecular mechanisms underlying the disease.

BNrich improves pathway enrichment analysis by incorporating Bayesian networks and gene expression data to prioritize impacted pathways.

This study investigates factors influencing gene length in the human genome, finding correlations with SNPs, introns, protein size, and tissue-specific expression, and linking gene length to functional roles in development and disease.

The paper introduces DIA-norm, a dynamic pathway analysis tool for time-course data, and shows it identifies more biologically relevant pathways than DAVID while demonstrating pigs as a better model for human adipogenesis than mice.

The paper introduces a method to reconstruct disease-associated protein-protein interaction networks using comprehensive influence analysis, outperforming existing methods and applying it to prostate cancer and Parkinson's disease.

This paper introduces CPGD, a novel algorithm for identifying personalized driver genes and combinational drugs for individual cancer patients, showing improved performance over existing methods.

The paper identifies top disease-associated cells and genes in IMIDs and proposes drug repurposing targets using a DIME network approach.

The paper introduces an R software package for differential network analysis to predict disease mechanisms, drug targets, and biomarkers using gene expression data from two conditions, validated on prostate cancer datasets.

The paper constructs a disease network based on protein domain connections, revealing new disease pairs and drug repurposing opportunities not captured by traditional gene-based methods.

This paper develops hybrid protein representations combining sequence and network data to improve tissue-specific function prediction, achieving superior performance across 13 tissues.

The paper introduces Biomarker2vec, a method that predicts multi-type relationships between biomarkers using a heterogeneous network and vector representation, achieving high performance on 18 relationships.