Bioinformatics and Genomic Networks

The paper predicts pathway involvement for all KEGG pathway and compound entries using machine learning, achieving higher performance than metabolic pathways alone.

The paper introduces CORNETO, a unified framework for knowledge-driven network inference from omics data, improving performance on biological tasks using simulated and real data.

The paper improves protein interaction prediction by addressing incomplete features and proteome coverage, identifying cell line-specific complexes.

The paper evaluates methods for inferring kinase-driven signaling pathways using phosphoproteomics data and kinase-substrate networks, finding that literature-curated networks with network propagation best recover ground truth interactions.

The paper introduces todenE, a clustering method combining topology and density-based approaches to group PAGs into Super-PAGs, improving functional representation in genomics.

This study benchmarks machine learning methods for predicting synthetic lethality in cancer using matrix factorization and deep learning models, highlighting performance differences and emphasizing data and scenario importance.

The paper uses mathematical models to show that intrinsic disorder in proteins correlates with clustering in eukaryotic interactomes and organism seniority.

E.PathDash provides a tool for analyzing gene expression data from pathogens, enabling rapid pathway activation analysis and reproducibility assessment across multiple studies.

mulea provides an R package for enrichment analysis across multiple ontologies with empirical FDR correction to address limitations in traditional gene set enrichment methods.

FunCoup 6 enhances functional association networks with directed links, expanded coverage, and improved tools for analyzing protein interactions across species.

PhenoMultiOmics visualizes multi-omics networks around enzymatic reactions in cancer, revealing disrupted pathways and biomarkers through curated databases and analysis modules.

SOFA identifies novel sources of variation in multi-omic data by incorporating known covariates and focusing on unexplained factors.

FishCODE provides a comprehensive multi-omics platform for fish research, offering extensive datasets and annotations to enhance data exploration and analysis.

The paper proposes a deep learning model to predict human essential genes using weighted PPI networks and achieves better performance than centrality-based methods.

The paper introduces a software tool for interpreting and visualizing KEGG pathway and Gene Ontology functional analysis, offering interactive networks, hierarchical analysis, and efficient processing.

GAUDI integrates multi-omics data using UMAP embeddings and density-based clustering to uncover non-linear relationships and identify interpretable latent factors across datasets.

This study uses scRNA-seq data with machine learning to predict plant metabolic pathway genes, finding higher accuracy compared to bulk RNA-seq and suggesting more benchmark genes improve predictions.

OmicsFootPrint integrates and interprets multi-omics data using circular images and deep neural networks, achieving high AUC scores in cancer subtype classification and drug response prediction.

This study develops a method to predict protein-protein direct interactions and stable complexes using co-expression and co-evolution data, achieving high AUROC scores and better performance than existing tools.

RummaGEO automatically mines gene sets from GEO RNA-seq studies for human and mouse, enabling signature search and metadata analysis.

This paper introduces a knowledge graph to prioritize cancer mutations for targeted therapy by integrating clinical, functional, and structural data.

The paper introduces HELP, a framework for labelling and predicting human essential genes, showing it outperforms existing methods in balancing false and true positive rates.

The paper estimates protein copy numbers from mRNA data across 369 cell lines and applies it to reclassify breast cancer subtypes and predict viral receptor abundances.

This paper introduces derivative profiling (DP) to identify differentially changing signals in dynamic omics data, showing it is statistically comparable to existing methods and computationally efficient.

The paper shows that gene loss in cancer is often compensated by paralog protein abundance changes, stabilizing protein-protein interaction networks and creating therapeutic vulnerabilities.

The paper customizes gene set collections for single-cell RNA-seq by incorporating cell type-specific gene expression data to improve gene set testing power and interpretability.

GNN4DM identifies overlapping functional disease modules by integrating network topology and genomic data, outperforming existing methods and discovering novel multimorbidity modules.

This study identifies shared differentially expressed genes between COVID-19 and gastric cancer, revealing immune-related pathways and potential therapeutic agents.

MONFIT integrates time-series single-cell RNA-sequencing with bulk proteomics and metabolomics to identify PD-associated genes and pathways.

BioPAX-Explorer simplifies querying and analyzing BioPAX biological pathway data by providing an object-oriented Python framework with query capabilities and tools for data extraction and workflow development.