Bioinformatics and Genomic Networks

The paper evaluates deep learning models for protein-protein interaction prediction, finding they rely on data leakage and sequence similarity, and advocates for baseline comparisons.

The paper introduces rROMA, a software package for computing gene set activity from omics data, and applies it to cystic fibrosis transcriptomic data to identify disease-related pathways and potential biases.

The paper curates 17 Drosophila signaling pathways using experimental evidence and integrates them into FlyBase with visualization tools.

NetVA identifies key molecules in disease-specific PPI networks using network vulnerability and EVC+ analysis, highlighting essential proteins and hubs in breast cancer.

The paper introduces IDSL.GOA, a tool for Gene Ontology analysis of metabolomic datasets, highlighting its ability to identify overrepresented GO terms not covered by existing pathway databases.

GenePlexusZoo improves gene classification by integrating molecular networks from multiple species into a shared feature space, enhancing both single-species and cross-species predictions.

FAVA infers high-quality functional protein interaction networks from scRNA-seq and proteomics data, outperforming existing methods and identifying interactions for understudied proteins.

mosGraphGen generates multi-omics signaling graphs to enable integrative and interpretable graph AI model development by converting multi-omics data into biologically meaningful graphs.

The paper introduces secCellFie, a tool that infers secretory and metabolic pathway activity from omic data, demonstrating its application in immune cells, NAFLD hepatokine secretion, and antibody production in Chinese Hamster Ovary cells.

This paper introduces an axes-based network embedding method to uncover functional mechanisms of cells by analyzing the structure of embedding spaces rather than gene clusters.

collectNET infers cell-cell communication networks using three methods and provides an online tool with visualization and a human communication atlas.

This paper identifies potential pain-related genes using machine learning and interaction networks, finding JAK2/STAT3, ErbB, and Rap1 pathways as promising targets.

The paper annotates disease variants in the context of biological pathways to infer molecular phenotypes and model variant impacts.

The paper introduces a network permutation framework to assess how node degree influences edge prediction in biomedical networks and provides an edge prior as a baseline for performance evaluation.

iTraNet is a web tool for visualizing and analyzing trans-omics networks, revealing differences in hub molecule responses to glucose between wild-type and ob/ob mice.

This paper introduces a faster approximation of the IsoRank method for aligning large biological networks, achieving comparable accuracy with significantly reduced computational time.

This paper develops Draphnet, a model that connects drug molecular effects to disease genetics to predict and explain drug impacts on disease phenotypes.

This paper introduces a non-parametric differential network analysis method to identify gene expression differences between males and females in diabetes datasets.

CAT Bridge identifies gene-metabolite associations using time-series multi-omics data and benchmarks statistical methods for longitudinal analysis.

The paper introduces MIPath, a pathway analysis method using mutual information to assess pathway-phenotype associations with fewer assumptions, showing better performance on single-cell datasets.

This study investigates how network depth affects gene set inference from bacterial transcriptomes using denoising autoencoders and identifies genes uniquely induced during human colonization.

The paper analyzes population-specific mutations' impact on the human interactome using in-silico edgetic profiling of nsSNVs from the 1,000 Genomes Project.

The paper identifies shared differentially expressed genes between type 2 diabetes and three cancers, and finds hub genes that may predict cancer risk in diabetic patients.

The paper identifies a bias in gene selection between -omics experiments and publication, introduces a tool to find understudied genes, and demonstrates its application in vertebrate aging research.

RedRibbon is a rank-rank hypergeometric overlap method for comparing gene and transcript expression signatures, showing accuracy and efficiency on large datasets.

Splitpea quantifies changes in protein interaction networks due to alternative splicing in cancer by analyzing domain-domain interactions and differential exon usage data from TCGA and GTEx.

This paper introduces PFOCR, a pathway database derived from published figures, and demonstrates its utility in cancer subtype and grade prediction.

The paper analyzes virus-host protein co-expression networks to understand the temporal dynamics and strategies of viral infection.

PathExpSurv integrates known biological pathways with neural networks to improve survival analysis in cancer, identifying disease-related genes through pathway expansion.

The paper develops a high-throughput AP-MS method to map a comprehensive yeast interactome, doubling the number of proteins and tripling reliable interactions.