Improving Gene-Set Enrichment Analysis of RNA-Seq Data with Small Replicates

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Gene-set enrichment analysis (GSEA) is crucial for interpreting transcriptome data from microarray and RNA-seq experiments.
• RNA-seq data often have limited replicates, necessitating gene-permuting GSEA methods.
• Standard gene-permuting GSEA methods produce false positives due to inherent inter-gene correlations within gene sets.

Purpose of the Study:

• To develop an improved GSEA method for RNA-seq data that controls false positives.
• To enhance the discriminatory power of gene-permuting GSEA by accounting for inter-gene correlations.
• To provide a robust tool for accurate pathway analysis in RNA-seq studies.

Main Methods:

• Developed a novel simulation method to generate RNA-seq read counts with inter-gene correlations within gene sets.
• Proposed a modified one-tailed GSEA incorporating absolute gene statistics.
• Compared the proposed method against existing RNA-seq enrichment analysis tools.

Main Results:

• The proposed GSEA method significantly improved false-positive control compared to existing methods.
• Incorporating absolute gene statistics enhanced the discriminatory ability for identifying deregulated pathways.
• Analysis of both simulated and real RNA-seq data validated the method's superior performance.

Conclusions:

• The enhanced GSEA method offers superior false-positive control and biological relevance for RNA-seq data analysis.
• This approach addresses limitations of standard GSEA when applied to RNA-seq datasets with small sample sizes.
• An R package, AbsFilterGSEA, is available for practical implementation.