GOREA: Unbiased Interpretation of Functional Enrichment

Functional enrichment analysis is essential for extracting biological meaning from gene expression data. Gene set enrichment analysis (GSEA) and over-representation analysis (ORA) are widely used approaches for this purpose. However, interpreting the large number of enriched gene ontology biological process (GOBP) terms remains challenging. Existing tools such as simplifyEnrichment often yield overly general and fragmented keywords, and they do not effectively utilize quantitative metrics such as normalized enrichment scores (NES) or gene overlap proportions, thereby limiting biological interpretation and prioritization. To address these issues, we developed GOREA, an improved tool for summarizing GOBP terms. GOREA improves upon simplifyEnrichment by integrating binary cut and hierarchical clustering, incorporating GOBP term hierarchy to define representative terms, and ranking clusters based on NES or gene overlap proportions. Using ComplexHeatmap R package, GOREA visualizes results as a heatmap accompanied by a panel of broad GOBP terms and representative terms for each cluster, providing both general and specific biological insights. Compared to simplifyEnrichment, GOREA yields more specific and interpretable clusters while significantly reducing computational time. GOREA effectively identified distinct biological processes in immune-related data and revealed substantial overlap between GOBP terms and cancer hallmark gene sets, demonstrating its applicability across diverse biological contexts. These findings suggest that GOREA provides a substantial improvement over existing approaches and offers a scalable and efficient framework for GSEA and ORA across diverse biological contexts.