Model-free unsupervised gene set screening based on information enrichment in expression profiles

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Unsupervised gene set screening methods are crucial for identifying functional gene sets from expression data.
• Current methods often rely on assumed statistical models, potentially missing informative gene sets with non-conforming expression profiles.

Purpose of the Study:

• To develop a model-free unsupervised gene set screening method to overcome limitations of existing approaches.
• To introduce Matrix Information Enrichment Analysis (MIEA) for enhanced discovery of gene sets from transcriptome data.

Main Methods:

• Matrix Information Enrichment Analysis (MIEA) was developed as a model-free approach.
• MIEA screens gene sets based on the information richness of their expression profiles, without assuming specific models.

Main Results:

• MIEA demonstrated superior performance in benchmark tests against other unsupervised gene set screening methods.
• The method successfully detected singular expression profiles missed by existing approaches.
• MIEA showed broad applicability and robust performance across simulated and real transcriptome data.

Conclusions:

• Matrix Information Enrichment Analysis (MIEA) is a broadly applicable tool for unsupervised gene set screening.
• MIEA enhances the mining of regulatory programs from transcriptome data by capturing diverse expression patterns.