DBGSA: a novel method of distance-based gene set analysis

Area of Science:

• Genomics
• Bioinformatics
• Systems Biology

Background:

• Gene set analysis (GSA) enhances information extraction from gene expression profiles compared to single-gene analysis.
• Existing GSA methods often fail to detect gene sets composed of numerous minor-effect genes.

Purpose of the Study:

• To develop a novel distance-based GSA method capable of identifying gene sets with minor-effect genes.
• To improve the detection of functional gene sets associated with complex traits.

Main Methods:

• Proposed a distance-based GSA approach measuring the difference between gene groups of distinct phenotypes.
• Estimated P-values using two permutation methods and applied multiple hypothesis testing adjustments.
• Validated the method on simulated and real gene expression datasets.

Main Results:

• The gene resampling-based permutation method proved more suitable for GSA.
• Centroid and average linkage statistical distances efficiently detected gene sets with minor-effect genes.
• The developed method demonstrated effectiveness in identifying significant functional gene sets.

Conclusions:

• The novel distance-based GSA method enhances the identification of functional gene sets, especially those with subtle genetic contributions.
• The method provides a valuable tool for understanding complex traits through gene expression analysis.
• Publicly available Perl and R packages (DBGSA) facilitate the application of this method.