GeneWalk identifies relevant gene functions for a biological context using network representation learning

Abstract

The primary bottleneck in high-throughput genomics experiments is identifying the most important genes and their relevant functions from a list of gene hits. Existing methods such as Gene Ontology (GO) enrichment analysis provide insight at the gene set level. For individual genes, GO annotations are static and biological context can only be added by manual literature searches . Here, we introduce GeneWalk ( github.com/churchmanlab/genewalk ), a method that identifies individual genes and their relevant functions under a particular experimental condition. After automatic assembly of an experiment-specific gene regulatory network, GeneWalk quantifies the similarity between vector representations of each gene and its GO annotations through representation learning, yielding annotation significance scores that reflect their functional relevance for the experimental context. We demonstrate the use of GeneWalk analysis of RNA-seq and nascent transcriptome (NET-seq) data from human cells and mouse brains, validating the methodology. By performing gene-and condition-specific functional analysis that converts a list of genes into data-driven hypotheses, GeneWalk accelerates the interpretation of high-throughput genetics experiments.

Clinico-histopathologic and single-nuclei RNA-sequencing insights into cardiac injury and microthrombi in critical COVID-19

“…We then used a network representation learning framework (42) to identify regulator genes--DE genes with the most connections to all other DE genes and to their gene ontologies.…”

High content genome-wide siRNA screen to investigate the coordination of cell size and RNA production

“…This resulted in ~200 clusters with diversity in both functional annotation and cellular phenotypes. One or two genes were then sampled from each of these clusters, with individual perturbations prioritised using GeneWalk 19 scores and phenotype strength. This panel was further supplemented with a selection of unperturbed and manually selected perturbations resulting in selection of ~237, 136, 66 putative up, down and non-hits.…”

An Early Stage Researcher's Primer on Systems Medicine Terminology

“… 242 , 243 These give an overview and comparison of the use of graph embedding methods in three important biomedical link prediction tasks: drug-disease association prediction, drug–drug interaction prediction, and protein–protein interaction prediction; and two node classification tasks: medical term semantic type classification and protein function prediction. 244 These identify relevant gene functions for a biological context using network representation learning with a neural networks-based graph embedding method. In a neuroscience context, a random walk-based graph embedding method is used for embedded vector representations of connectomes to map higher-order relations between brain structure and function.…”