BIGwas: Single-command quality control and association testing for multi-cohort and biobank-scale GWAS/PheWAS data

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) and phenome-wide association studies (PheWAS) involving large sample sizes (1 million individuals) from biobanks present significant computational challenges.
• Current methods require substantial time, personnel, and computational resources, with a lack of automated workflow solutions for processing such massive datasets.

Purpose of the Study:

• To present BIGwas, a portable and fully automated pipeline for quality control and association testing of large-scale GWAS data from biobanks.
• To provide a user-friendly solution for researchers, regardless of their bioinformatics expertise or available computational resources.

Main Methods:

• Utilized Nextflow workflow and Singularity software container technology for portability and reproducibility.
• Implemented a fully automated pipeline for quality control and association testing of GWAS data.
• Developed a dynamic parallelization approach for efficient analysis on high-performance compute (HPC) systems.

Main Results:

• BIGwas successfully performs automated quality control and association testing for large-scale GWAS data.
• A single-command GWAS analysis of 974,818 individuals and 92 million genetic markers was completed in approximately 16 days on a small HPC system (7 compute nodes).
• The pipeline is resource-efficient, reproducible, and scalable for local computers or HPC environments.

Conclusions:

• BIGwas empowers researchers with limited bioinformatics knowledge and computational resources to conduct large-scale, multi-cohort GWAS.
• The pipeline facilitates the creation of custom genome-wide phenome-wide association study (PheWAS) resources.
• BIGwas is freely available, promoting wider adoption and accessibility in genetic research.