Bioinformatics and Genomic Networks

XomicsDB curates and provides paired multi-omics datasets for model organisms to facilitate data integration and systems biology research.

multiDEGGs performs differential network analysis across multi-omic data to identify biomarkers and improve predictive modeling in rheumatoid arthritis treatment response.

ppIRIS predicts bacterial protein-protein interactions using deep learning, achieving high accuracy and enabling proteome-wide screening, with applications in identifying virulence pathways and host-pathogen interactions.

This study quantifies genetic interaction conservation between two yeast species, finding significant conservation among functionally related genes but reduced conservation due to gene duplication.

This study evaluates integration strategies for multimodal bioinformatics data with structured missingness, finding late integration and logistic regression models perform best in breast cancer subtyping.

MORGaN is a self-supervised framework for drug target discovery that outperforms existing models by leveraging multi-omic, multi-relation gene networks and provides interpretable results.

The paper constructs an atlas of human multi-protein complexes by integrating computational structure predictions with genetic dependency data, identifying 354 high-confidence complexes including novel ones.

The paper introduces PIS, a method to improve DEG selection by optimizing thresholds for pathway enrichment, showing better recovery of true DEGs and stability across datasets.

HyWinNet predicts protein-protein interactions by integrating hyperbolic geometry and wavelet transforms to model multi-scale biological networks, outperforming existing methods in recovering known interactions and predicting functional sites.

A-LAVA identifies germline variants affecting metabolic pathways in cancer genomes through SNP, gene, and pathway-level analysis, improving accuracy by correcting for overlapping gene sets and reducing false positives.

EnrichMET provides an R package for fast and reproducible pathway enrichment analysis in metabolomics, supporting multiple methods and offering both command-line and Shiny interfaces.

MAPA uses large language models to reduce redundancy in pathway analysis and improve biological interpretation of multi-omics data.

The paper constructs a molecular interaction map of colorectal cancer, identifies disease signatures, and predicts therapeutic targets through simulation and validation.

The paper analyzes cell-cell interactions (CCIs) in transcriptomic data, showing they are more co-expressed in cancer and associated with patient survival and immunotherapy response.

SEAHORSE systematically explores associations across heterogeneous omics data from GTEx and TCGA, revealing unexpected links between height and cancer-related pathways, and age with WNT signaling.

This study evaluates LLMs' ability to infer biochemical reactions and disease associations from protein pathways, finding poor performance and introducing a benchmark dataset.

This paper introduces AIGE, an AI tool that automates functional gene annotation with high accuracy and self-assessment capabilities.

The paper constructs a comprehensive signaling pathway network for hepatocellular carcinoma and identifies key regulators PTK2, GSK3β, and β-catenin as critical for cancer progression.

DirectContacts2 identifies direct protein interactions using mass spectrometry data and machine learning, improving accuracy over existing resources and providing structural models for disease-related complexes.

The paper evaluates genome-scale protein function annotations for completeness, coherence, and consistency, finding that computational methods fail to produce biologically plausible annotations despite high individual protein accuracy.

The paper constructs a genome-scale genetic interaction network from 4 million double mutants in HAP1 cells, revealing conserved genetic network principles across species and uncovering novel gene functions and cancer dependencies.

The paper finds that positive genetic interactions have greater scientific impact but are underrepresented in literature, despite equal detection capability.

GhostBuster predicts gene functions and annotations while minimizing literature bias, outperforming biased datasets in identifying novel gene annotations.

Proteomizer infers proteomic profiles from transcriptomic and miRNomic data, improving differential expression analysis accuracy but with limited generalizability.

This paper introduces Regulation Flow Analysis (RFA), a method that generates causal hypotheses of drug mechanisms using knowledge graphs, improving understanding of biological processes and drug development.

D-CoMEx identifies phenotype-specific miRNA biomarkers linked to Parkinson's disease and Breast Cancer by integrating co-expression networks with differential analysis.

The paper introduces iTissue, a method integrating spatial and single-cell transcriptomics with molecular networks to identify metastasis-related genes in ccRCC.

CladeOScope-GSA predicts functional gene interactions using phylogenetic profiling across 1905 species, offering enhanced analysis and visualization tools for comparative genomics.

The paper introduces a circular van Krevelen diagram for automated metabolic pathway visualization, improving consistency for data interpretation and map creation.

The paper introduces a gene set foundation model trained on diverse gene sets to predict gene functions, achieving superior performance compared to existing methods.