Bioinformatics and Genomic Networks

OMICON: a community resource for studying gene coexpression networks in normal and neoplastic human brain samples

OMICON provides a resource for studying gene coexpression networks in human brain samples, offering standardized data and tools for analysis and discovery.

BfBio: a graph-based tool for the prediction of Angiogenic Stalk Cell genes using a Personalized PageRank algorithm

BfBio predicts angiogenic stalk cell genes using a Personalized PageRank algorithm, achieving high AUC-ROC performance and identifying novel cancer-related genes.

BLOBFISH: Bipartite Limited Subnetworks from Multiple Observations using Breadth-First Search with Constrained Hops

BLOBFISH identifies gene regulatory subnetworks connecting predefined gene sets using breadth-first search on bipartite graphs, applied to GTEx data to explore tissue-specific relationships.

OmicsFM brings proteomics into the foundation model era

OmicsFM is a proteomics foundation model that matches transcriptomic models in performance, revealing complementary biology between proteomics and transcriptomics.

Finetuning Foundation Models for Temporal Clinical Transcriptomics Data

The paper finetunes gene embeddings from foundation models for temporal clinical transcriptomics data, improving gene network analysis and detection of disease progression and drug response mechanisms.

Benchmarking Graph Neural Networks for Multi-Omics Cancer Subtyping using Methylation and Gene Expression Profiles

This study benchmarks graph neural networks for cancer subtyping using methylation and gene expression data, finding that ChebNet performs best and simple readout layers improve performance.

Degree-ranked gene lists omit the cross-module connectors, and a partition-free centrality recovers them

The paper shows that standard gene centrality measures miss cross-module connectors and introduces EDVS, a partition-free centrality measure that recovers these genes without functional annotations.

EnSEMBLE: a framework for enhancer-anchored pathway analysis that locks in enhancer-corroborated pathways from transcriptome sequencing data for biological validation

EnSEMBLE reduces redundant gene sets to enhancer-corroborated pathways by integrating enhancer RNA data with pathway enrichment analysis, improving biological relevance and reducing false positives.

When the Background Matters: Topic-Dependent reference lists in GWAS and Exome Analyses

The paper shows that using mismatched genome or exome backgrounds in GO analysis affects significance and introduces annotation bias, with practical implications for GWAS and WES studies.

Port of Protein-Protein Interactomes: An experiment-based protein-protein interactome database for rice

The paper introduces POPPIN, an experimentally verified PPI database for rice, containing over 150,451 PPI-related entries and functional annotations.

Signature Recontextualization: Mapping perturbational signatures across biological contexts

This paper introduces a benchmarking framework for predicting perturbation signatures across biological contexts, showing that projection and network-based methods perform well across diverse datasets and contexts.

PINT: Pathway-pathway interactions for predicting interpretable clinical outcomes from gene expression

PINT predicts clinical outcomes from gene expression by modeling pathway-pathway interactions, outperforming benchmarks and revealing biologically meaningful pathway relationships.

Integrated metabolomics data analysis to generate mechanistic hypotheses with MetaProViz

The paper introduces MetaProViz, an R package for metabolomics data analysis that integrates prior knowledge to generate mechanistic hypotheses, using kidney cancer data to demonstrate its capabilities.

Exploring vulnerable proteins in the progression of head and neck squamous cell carcinoma

The paper identifies vulnerable proteins in HPV-positive and HPV-negative HNSCC using PPI networks and multi-omics data, and predicts synthetic lethal partners with potential therapeutic targets.

Improved prediction of virus-human protein-protein interactions by incorporating network topology and viral molecular mimicry

This paper introduces vhPPIpred, a machine learning method for predicting virus-human PPIs that outperforms existing methods and provides a benchmark dataset.

RingNet: An Interactive Platform for Multi-Modal Data Visualization in Networks

RingNet is a web-based tool for interactive visualization of multi-modal biological networks, enabling pattern identification and cross-modal relationship discovery without requiring advanced programming skills.

CompBio and MIRaS : A Multi-omic Analysis Platform Built on a Memory-Based Intelligence Engine

The paper introduces CompBio, a multi-omic analysis platform using MIRaS, a memory-based intelligence engine, to analyze omics data without relying on human-curated knowledgebases, producing traceable and statistically significant results.

Sample-specific protein-protein interaction networks inferred from transcriptomics and proteomics show high similarities

The paper compares transcriptomic and proteomic estimates of sample-specific protein-protein interaction networks, finding similar size-adjusted characteristics and time-related trends but differing rates of change.

pysigscore: gene signatures scoring across bulk and single-cell transcriptomics

pysigscore scores gene signatures in bulk and single-cell RNA-seq data, validating its performance on multiple datasets.

A Framework for Benchmarking Pathway Reconstruction Algorithms

This paper benchmarks 14 pathway reconstruction algorithms across 822 datasets using a standardized framework to evaluate performance, similarity, and computational efficiency.

A map of human protein-protein interaction embeddings for functional discovery

The paper introduces MAPPIE, a method to map human protein-protein interactions into a 2D space for functional discovery, outperforming baselines in sparse interaction regions.

Inferring disruption of directed graphs using LIKA reveals altered protein phosphorylation networks in schizophrenia

LIKA, a likelihood model over many-to-many kinase-substrate networks, beats KSEA in simulations with shared substrates, recovers EGFR and ABL1 in cell lines, and ranks MAPK3 first in schizophrenia cortex.

LagCI Enables Inference of Temporal Causal Relationships from Dense Multi-Omic Time Series

LagCI identifies temporal causal relationships in dense multi-omics time series by analyzing lag-correlation profiles and filtering spurious associations.

Mango: Unearthing Patterns in Large-Scale Biological Data Through Interactive Correlation Analysis

Mango enables interactive correlation analysis of tabular biological data to distinguish biases from causal relationships in genomic and transcriptomic datasets.

KlinkPPI: Single Point of Access to Protein-Protein Interactions Across Databases

KlinkPPI provides a unified interface to retrieve and compare protein-protein interaction data from multiple databases without merging or consensus.

MEGA-ODE: Learning Biologically Structured and Navigable Continuous Perturbation Dynamics from Sparse Omics

MEGA-ODE reconstructs dynamic omics landscapes from sparse data, predicts unmeasured states, and prioritizes perturbations toward biological endpoints.

GeneInsight: Condensing Gene Set Knowledge via Language Models

GeneInsight synthesizes gene set annotations using topic modeling and language models to provide interpretable summaries that reveal biologically significant insights overlooked by traditional enrichment analyses.

Confidence-aware learning for transcriptome-based prediction of OXPHOS genes in Caenorhabditis elegans under incomplete functional annotation

The paper develops a confidence-aware learning framework to predict OXPHOS genes in C. elegans using transcriptomic data and experimental validation.

Explainable HGT-based framework for predicting human dark kinase protein-pathway associations by leveraging BERT-based embeddings and WGAN-GP

The paper presents a framework using BERT embeddings and WGAN-GP to predict pathway associations for human dark kinase proteins, achieving high accuracy in classification and pathway clustering.

Genome-scale prediction of context-specific synthetic lethality beyond protein interaction networks

This paper introduces SLxGO, a network-independent machine learning framework for predicting context-specific synthetic lethal gene pairs, achieving superior performance and validating interactions experimentally.