GeneSetCluster: a tool for summarizing and integrating gene-set analysis results

BMC Bioinformatics

Ewoud Ewing¹  Maja Jagodic¹  Nuria Planell-Picola²  David Gomez-Cabrero³ 

[1] Department of Clinical Neuroscience, Center for Molecular Medicine, Karolinska Institutet, 171 77, Stockholm, Sweden;Translational Bioinformatics Unit, Navarrabiomed, Complejo Hospitalario de Navarra (CHN), Universidad Pública de Navarra (UPNA), IdiSNA, Pamplona, Spain;Translational Bioinformatics Unit, Navarrabiomed, Complejo Hospitalario de Navarra (CHN), Universidad Pública de Navarra (UPNA), IdiSNA, Pamplona, Spain;Unit of Computational Medicine, Department of Medicine, Solna, Center for Molecular Medicine, Karolinska Institutet, 171 77, Stockholm, Sweden;

关键词: Data-mining;    Gene-set enrichment;    Clustering pathways;    Overlapping pathways;    Clustering gene-sets;   

DOI  :  10.1186/s12859-020-03784-z

来源: Springer

【 摘 要 】

BackgroundGene-set analysis tools, which make use of curated sets of molecules grouped based on their shared functions, aim to identify which gene-sets are over-represented in the set of features that have been associated with a given trait of interest. Such tools are frequently used in gene-centric approaches derived from RNA-sequencing or microarrays such as Ingenuity or GSEA, but they have also been adapted for interval-based analysis derived from DNA methylation or ChIP/ATAC-sequencing. Gene-set analysis tools return, as a result, a list of significant gene-sets. However, while these results are useful for the researcher in the identification of major biological insights, they may be complex to interpret because many gene-sets have largely overlapping gene contents. Additionally, in many cases the result of gene-set analysis consists of a large number of gene-sets making it complicated to identify the major biological insights.ResultsWe present GeneSetCluster, a novel approach which allows clustering of identified gene-sets, from one or multiple experiments and/or tools, based on shared genes. GeneSetCluster calculates a distance score based on overlapping gene content, which is then used to cluster them together and as a result, GeneSetCluster identifies groups of gene-sets with similar gene-set definitions (i.e. gene content). These groups of gene-sets can aid the researcher to focus on such groups for biological interpretations.ConclusionsGeneSetCluster is a novel approach for grouping together post gene-set analysis results based on overlapping gene content. GeneSetCluster is implemented as a package in R. The package and the vignette can be downloaded at https://github.com/TranslationalBioinformaticsUnit

【 授权许可】

CC BY