Methods for multiple outcome meta-analysis of gene-expression data.

Area of Science:

• Genomics
• Biostatistics
• Bioinformatics

Background:

• Meta-analysis synthesizes evidence from diverse studies, including high-throughput genomics.
• Multivariate meta-analysis offers advantages over univariate methods when dealing with multiple outcomes or treatments.
• Gene expression studies are crucial for understanding complex traits and diseases.

Purpose of the Study:

• To adapt and apply the multivariate meta-analysis method to gene expression data.
• To identify genes with pleiotropic effects, influencing multiple disease outcomes.
• To discover differentially expressed genes (DEGs) not detectable by univariate approaches.

Main Methods:

• Application of a multivariate meta-analysis framework to gene expression datasets.
• Joint modeling of multiple outcomes to increase statistical power and enable global tests.
• Testing the method on inflammatory bowel disease (IBD) data, including Crohn's disease (CD) and Ulcerative colitis (UC).

Main Results:

• The multivariate meta-analysis successfully identified genes with pleiotropic effects.
• The method revealed genes with expression changes not apparent in univariate analyses.
• Demonstrated utility in complex trait analysis, specifically for IBD subtypes.

Conclusions:

• Multivariate meta-analysis is a powerful tool for gene expression data analysis.
• This approach enhances the detection of pleiotropy and DEGs in complex diseases.
• The developed method and Stata code provide valuable resources for genomic research.