MAAT

MAAT expands transcriptome-wide association studies by incorporating multiple functional annotations through a nonparametric Bayesian framework and generates a metric indicating the most important annotation.

Key Features:

• Nonparametric Bayesian Framework: MAAT employs a nonparametric Bayesian approach for transcriptome-wide association studies that does not assume a positive correlation between annotation scores and SNPs' effect sizes.
• Multiple Annotation Integration: MAAT incorporates multiple functional annotations beyond predominantly epigenomic annotations into TWAS analysis.
• Annotation Importance Metric: MAAT generates a new metric that indicates the most important annotation among those included in the model.

Scientific Applications:

• Post-GWAS Gene Regulation Discovery: MAAT translates genome-wide association study variants into regulated genes in the post-GWAS era.
• Annotation-Assisted TWAS: MAAT enhances statistical power in transcriptome-wide association studies by integrating diverse functional annotation information.

Methodology:

MAAT uses a nonparametric Bayesian framework to model multiple functional annotations in transcriptome-wide association studies without assuming positive correlation between annotation scores and SNPs' effect sizes.