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

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene-set analysis tools identify biological insights by detecting over-represented gene-sets associated with specific traits.
• Current tools, including those for RNA-sequencing and DNA methylation analysis, often yield numerous gene-sets with overlapping content, complicating interpretation.
• Interpreting large numbers of gene-sets with similar gene compositions poses a significant challenge for researchers seeking clear biological understanding.

Purpose of the Study:

• To introduce GeneSetCluster, a novel computational approach for clustering gene-set analysis results.
• To address the complexity of interpreting gene-set analysis outputs by grouping gene-sets with shared gene content.
• To facilitate biological interpretation by organizing numerous gene-sets into coherent, manageable clusters.

Main Methods:

• GeneSetCluster employs a distance score based on the overlap of gene content between gene-sets.
• The approach clusters gene-sets from single or multiple experiments and analysis tools.
• Utilizes a gene-overlap metric to group gene-sets with similar functional definitions.

Main Results:

• GeneSetCluster successfully groups identified gene-sets based on their shared gene composition.
• The clustering identifies distinct groups of gene-sets with similar underlying gene content.
• This facilitates a more focused and efficient interpretation of complex gene-set analysis results.

Conclusions:

• GeneSetCluster offers a novel method for organizing and interpreting post-gene-set analysis results.
• The R package implementation provides a practical tool for researchers working with gene-set data.
• Available on GitHub, GeneSetCluster enhances the ability to extract meaningful biological insights from complex genomic datasets.