Bioinformatics and Genomic Networks

The paper introduces a method using Krylov subspace trajectories for analyzing biological networks, demonstrating its application in community detection and perturbation analysis on the C. elegans neural network.

This paper introduces PerturbGraph, a graph neural network framework that predicts transcriptional responses to unseen gene perturbations by integrating biological priors, outperforming existing methods on benchmark datasets.

This paper introduces a graph-based machine learning approach to predict gene perturbation effects on molecular phenotypes using existing experimental data and biological networks.

PACMON integrates multi-omics data from perturbation screens to infer pathway-level responses to experimental perturbations, achieving high accuracy and scalability.

This paper introduces REGEN, a GNN framework for learning gene interaction networks from bulk transcriptomic data and predicting patient outcomes, showing improved performance across cancer types.

This paper introduces differentiable GSEA (dGSEA) to improve pathway-level interpretation in transcriptomic learning by aligning training objectives with downstream analysis.

The paper expands TheCellMap.org to include human genetic interactions from CRISPR-based analysis in HAP1 cells, enabling visualization and exploration of the network.

EBEx prioritizes disease-relevant genes from transcriptomic data in COPD by integrating data-driven signals with curated gene sets and network-based expansion, identifying biologically relevant candidates and molecular subgroups.

This study prioritizes candidate genes in Alzheimer's disease using a network-based framework that integrates gene co-expression analysis with centrality measures to identify biologically meaningful genes.

MESSI benchmarks multimodal integration methods using standardized workflows and cross-validation, showing method performance varies by dataset and task.

GNExT is a web-based platform that integrates GWAS data with systems medicine approaches to identify potential drug repurposing candidates and enable large-scale network medicine investigations.

CAPRINI-M curates cardiac protein interactions using AI, integrating literature and structural data to improve PPI resource quality and performance.

miREA improves miRNA pathway analysis by incorporating miRNA-gene interactions, outperforming node-based methods in sensitivity and biological interpretability across cancer datasets.

This study identifies common molecular pathways and hub genes linking Alcohol Use Disorder and psychiatric disorders through integrated transcriptome analysis and network-based methods.

This paper analyzes single-sample network inference methods, identifying trade-offs between accuracy and specificity and suggesting opportunities for methodological synergy.

This study identifies shared molecular signatures across PTSD, schizophrenia, and bipolar disorder using transcriptomic data and machine learning, highlighting HLA-DRA as a potential biomarker.

This paper predicts synthetic lethal gene pairs using a foundation model-based framework with uncertainty calibration, improving performance without relying on PPI or GO data.

The paper develops an integrated similarity score for gene expressions using functional annotation to improve gene similarity and classify unclassified yeast genes.

TSUMUGI identifies gene networks driven by phenotypes using IMPC knockout mouse data to facilitate hypothesis generation for complex phenotypes.

The paper introduces a graph-based framework that integrates statistical distributions into omics data to improve predictive performance and biological interpretability in cancer datasets.

LLM-PathwayCurator converts pathway enrichment results into auditable claims with quality assurance, showing variable coverage across cancer cohorts.

The paper introduces Mahi, a graph neural network framework that integrates molecular context and functional connectivity to predict gene essentiality and perturbation responses across tissues.

datadrivenhypothesis.org integrates gene and pathway data to uncover metabolic connections and generate hypotheses about gene function.

thematicGO organizes GO enrichment terms into biological themes to reduce redundancy and improve readability in transcriptomic studies.

MultiGEOmics integrates multi-omics data using graph-based methods to model cross-omics regulatory signals and handle missing data, showing strong performance across cancer and Alzheimer's datasets.

EMTscore provides an efficient tool for unbiased scoring of EMT pathways in single-cell or bulk omics data and enables rapid screening of gene set correlations.

PAGAN predicts digenic interactions by learning network representations of gene pairs from single-gene data in biological networks.

This paper develops a deep-learning method to unify genetic evidence from common and rare diseases by integrating diverse data sources and identifying shared genetic mechanisms.

The paper uses Bayesian network analysis to identify gene modules in Nicotiana benthamiana during interfamily grafting, revealing six temporal expression patterns and functional enrichments.