PGS: a tool for association study of high-dimensional microRNA expression data with repeated measures

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• MicroRNAs (miRNAs) are key gene expression regulators.
• High-dimensional miRNA data often involves repeated measures over time.
• Existing methods like site-by-site (SBS) testing may not fully leverage longitudinal data.

Purpose of the Study:

• To develop a novel analytical framework for high-dimensional miRNA expression data with repeated measures.
• To address the limitations of univariate and SBS testing in longitudinal miRNA studies.
• To improve the power and biological interpretability of miRNA association analyses.

Main Methods:

• A penalized regression model incorporating a grid search method (PGS) was developed.
• The PGS method was applied to analyze a real-world miRNA dataset.
• Performance was compared against traditional SBS testing using simulations and real data.

Main Results:

• PGS demonstrated smaller phenotype prediction errors compared to SBS testing.
• PGS identified a higher enrichment of phenotype-related biological pathways.
• Simulations confirmed PGS provides more accurate estimates and higher sensitivity with comparable specificity.

Conclusions:

• The proposed PGS method offers a powerful approach for analyzing longitudinal high-dimensional miRNA expression data.
• PGS enhances biological insights by better utilizing the temporal nature of the data.
• This framework provides a more robust and informative alternative to conventional statistical methods.