Combining multiple tools outperforms individual methods in gene set enrichment analyses

Abstract

MotivationGene set enrichment (GSE) analysis allows researchers to efficiently extract biological insight from long lists of differentially expressed genes by interrogating them at a systems level. In recent years, there has been a proliferation of GSE analysis methods and hence it has become increasingly difficult for researchers to select an optimal GSE tool based on their particular dataset. Moreover, the majority of GSE analysis methods do not allow researchers to simultaneously compare gene set level results between multiple experimental conditions.ResultsThe ensemble of genes set enrichment analyses (EGSEA) is a method developed for RNA-sequencing data that combines results from twelve algorithms and calculates collective gene set scores to improve the biological relevance of the highest ranked gene sets. EGSEA’s gene set database contains around 25 000 gene sets from sixteen collections. It has multiple visualization capabilities that allow researchers to view gene sets at various levels of granularity. EGSEA has been tested on simulated data and on a number of human and mouse datasets and, based on biologists’ feedback, consistently outperforms the individual tools that have been combined. Our evaluation demonstrates the superiority of the ensemble approach for GSE analysis, and its utility to effectively and efficiently extrapolate biological functions and potential involvement in disease processes from lists of differentially regulated genes.Availability and ImplementationEGSEA is available as an R package at http://www.bioconductor.org/packages/EGSEA/. The gene sets collections are available in the R package EGSEAdata from http://www.bioconductor.org/packages/EGSEAdata/.Supplementary information Supplementary data are available at Bioinformatics online.

Rotenone exerts developmental neurotoxicity in a human brain spheroid model

“…The single-cell RNA-Seq data from these studies represent an invaluable source to define gene sets made of enriched/specific markers of cells found in these organoids and in the two surveyed brain areas. EGSEA (Alhamdoosh et al, 2017) was used to find the closest matches to the differentially expressed genes found in BrainSpheres at the three time-points, using a significance score (from zero to 100, with zero being the least significant and 100 representing the most significant gene sets. Heatmap of EGSEA for 123 gene sets obtained from the single-cell RNA-Seq datasets showed high scores at the three time-points for the three following gene sets: xiang_mgeo_72days_8 (minimum score = 97, p.adj = 2e-20), quadrato_6months_14 (minimum score = 87, p.adj = 2e-14, and sloan_130days_1 (minimum score = 88, p.adj = 2e-20) (Fig.…”

HIF‐1 Signaling Pathway Implicated in Phenotypic Instability in a Chinese Hamster Ovary Production Cell Line

“…GSEA, using Kyoto Encyclopedia of Genes and Genomes (KEGG) as the reference database, was employed to provide biological pathway relevance to the DEA analysis. To avoid methodology associated biases, a total of eight GSEA algorithms ( ora , camera , roast , fry , plage , zscore , gsva , ssgsea ) were integrated using Bioconductor's Ensemble Gene Set Testing workflow . This workflow utilizes the linear model generated by limma as input, which includes the β‐coefficient and p ‐value of the cell:generation interaction term of every gene in the count matrix.…”

“…To further increase the statistical robustness, the Benjamini–Hochberg false discovery rate (FDR) procedure was performed on the obtained P avg (γ i ) to correct any multiple testing related type‐I errors. In addition, to illustrate the overall magnitude of changes in gene expression for a given gene set γ i , an average log FC value was calculated as , where the log FC j is the log2 of the fold‐change of the j th gene in γ i . The log FC j equals the β‐coefficient of the interaction term between Cell Line A and generation in the limma linear regression model and reflects the change of gene j for its expression level in Cell Line A versus B over the generational time course that was studied.…”

Phytoplankton transcriptomic and physiological responses to fixed nitrogen in the California current system

“…Multiple approaches were applied to PLOS ONE examine responses at the phylogroup and individual strain levels and also due to variability among some treatment replicates (Table 1). We used an Ensemble of Gene Set Enrichment Analyses (EGSEA; [62]) to identify sets of genes that collectively were significantly differentially expressed based on a consensus of twelve GSEA algorithms (Benjamini-Hochberg adjusted p-value < 0.01, unadjusted p-value calculated using Wilkinson's method [64] to combine p-values from the GSEA algorithms). We defined each gene set to contain MicroTOOLs gene targets from a specific phylogroup and physiological response, e.g.…”