Computationally efficient meta-analysis of gene-based tests using summary statistics in large-scale genetic studies

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genetics

Background:

• Meta-analysis of gene-based tests is crucial for genetic association studies.
• Current methods necessitate sharing large covariance matrices, posing challenges for large-scale studies with multiple phenotypes.

Purpose of the Study:

• To introduce REMETA, an efficient tool for gene-based meta-analysis.
• To overcome the data sharing limitations of existing meta-analysis approaches.

Main Methods:

• REMETA utilizes a single sparse covariance reference file per study, rescaled for each phenotype using single-variant summary statistics.
• New methods were developed for binary traits with case-control imbalance and for estimating allele frequencies, genotype counts, and effect sizes for burden tests.

Main Results:

• Demonstrated REMETA's performance and advantages through a meta-analysis of five traits in 469,376 UK Biobank samples.
• The approach effectively handles large datasets and diverse phenotypes.

Conclusions:

• REMETA offers an efficient solution for gene-based meta-analysis, reducing computational and storage burdens.
• The open-source software facilitates meta-analysis across large-scale exome sequencing studies, promoting wider research collaboration.