Bioinformatics and Genomic Networks

This paper introduces a statistical physics approach using the random-field Ising model to detect disease modules, showing improved efficiency and robustness compared to existing methods.

The paper presents a LINCS Microenvironment perturbation dataset cataloging transcriptional, proteomic, epigenomic, and phenotypic responses of MCF10A cells to six ligands, enabling integrative analysis of molecular and phenotypic changes.

This paper introduces a unified framework for gene set analysis that improves performance and interpretability by comparing effect size distributions between gene sets and subgroups.

DGMP identifies cancer driver genes using a combination of DGCN and MLP on multi-omics data, outperforming existing methods by leveraging directed GRN structure and gene features.

BIONIC integrates biological networks using deep learning to improve accuracy and scalability, identifying novel gene interactions in yeast.

The paper presents an integrated experimental and computational strategy to model protein-protein interactions in cellular context, applied to the TNF-receptor signaling complex.

The paper constructs a gene feature dataset for Arabidopsis thaliana and develops a machine learning workflow to identify linked features, visualized as a Feature Important Network (FIN) to reveal novel biological insights.

The paper measures affinities of 65,000 PDZ domain interactions using quantitative fragmentomics, revealing viral interference hotspots and system-wide impacts of interactome hijacking.

This paper uses machine learning to integrate proteome and metabolome data, revealing new protein-metabolite connections and gene functions in yeast.

This study compares ranking aggregation methods for gene lists using simulated and real data to evaluate their performance in meta-analyses.

The paper introduces Bioteque, a large-scale biomedical knowledge graph resource that provides pre-calculated descriptors for downstream machine learning tasks.

GraphBio is a web app that provides visualization tools for omics data, enabling non-programmers to generate publication-ready figures.

SUBATOMIC clusters multi-omics networks into interpretable modules, identifies hypoxia-related modules in cancer, and provides functional insights for unannotated genes.

This study uses machine learning to predict genes associated with RNA methylation pathways in humans using multi-omics data from the Harmonizome database.

This paper introduces GNN-SubNet, an explainable graph neural network framework for detecting disease subnetworks using patient-specific protein-protein interaction networks enriched with molecular data.

DrDimont predicts differential drug response by integrating multi-omics networks and provides explainable results, outperforming methods based on differential protein expression or PageRank in breast cancer data.

The paper constructs a tissue-level phenome-wide network map linking genes and phenotypes using colocalized signals from the UK Biobank and Genotype-Tissue Expression project.

DSCN selects effective target combinations by integrating CRISPR screening data and protein-protein interaction networks, showing high accuracy and speed compared to existing methods.

This paper introduces SCALES, a method that extracts hierarchical protein interaction networks from bacterial proteome variation patterns, revealing new genetic factors in Pseudomonas aeruginosa motility.

The paper identifies a core cellular network in fly brains using co-expression analysis across cell types and validates it in additional datasets.

This study uses an unsupervised machine learning approach to analyze pathways in COVID-19, identifying key pathways related to SARS-CoV-2 infection.

This study constructs protein-protein interaction networks for human bromodomain-containing proteins to analyze their functional roles and network properties.

iLINCS provides a web-based platform for analyzing omics signatures of diseases and cellular perturbations through integrative analytics and visualization tools.

The paper introduces ssNet, a method to integrate probabilistic functional networks without an external Gold Standard, showing comparable performance using Saccharomyces cerevisiae data.

The paper develops a method to analyze longitudinal multi-omics profiles of prediabetic and diabetic individuals to identify individual and group-level immune response patterns.

The paper identifies highly perturbed and hub genes associated with type 2 diabetes across human tissues using a bioinformatics workflow and meta-analysis.

cageminer integrates GWAS and gene coexpression networks to prioritize candidate genes by filtering for physical proximity to SNPs, coexpression with known trait-associated genes, and expression changes in relevant conditions.

RAPPPID improves protein-protein interaction prediction by using regularised AWD-LSTM twin networks, outperforming existing methods on unseen proteins.

The paper reanalyzes proteomics datasets from mouse and rat tissues to compare baseline protein expression across species and organs, finding high correlation among orthologs in brain, kidney, heart, and liver.

This work analyzes transcriptomic data across multiple biological contexts to identify unique and shared gene expression patterns using co-expression networks and pathway analysis.