simplifyEnrichment: A Bioconductor Package for Clustering and Visualizing Functional Enrichment Results

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Functional enrichment analysis (FEA) and gene set enrichment analysis (GSEA) are crucial bioinformatics tools for assessing gene list significance.
• Current methods for simplifying FEA/GSEA results often struggle with inter-cluster redundancy or inconsistent intra-cluster term similarity.

Purpose of the Study:

• To introduce a novel clustering method, binary cut, for functional term similarity matrices.
• To address the limitations of existing tools in summarizing and reducing redundancy in enrichment analysis outputs.

Main Methods:

• Developed and applied the binary cut algorithm for clustering functional term similarity matrices.
• Evaluated performance using simulated and real-world biological datasets.
• Compared clustering results using different similarity measures, including semantic similarity and gene overlap.

Main Results:

• Binary cut efficiently clusters functional terms into groups with consistent intra-group similarity and inter-group exclusivity.
• Semantic similarity measures demonstrated superior performance with the binary cut algorithm compared to gene overlap-based similarities.
• The R package simplifyEnrichment was developed to implement and visualize binary cut clustering.

Conclusions:

• Binary cut offers an effective solution for simplifying complex functional enrichment results.
• The simplifyEnrichment package provides a user-friendly tool for enhanced analysis and visualization of gene set enrichment data.
• This method improves the interpretability of large gene lists in biological research.