RápidoPGS: a rapid polygenic score calculator for summary GWAS data without a test dataset

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Polygenic scores (PGS) are used for genetic prediction of complex traits.
• Current PGS methods often require independent test datasets for parameter tuning, which can be computationally intensive.
• Alternative methods that tune parameters on training data are also computationally demanding.

Purpose of the Study:

• To develop a fast and computationally efficient method for calculating PGS.
• To present RápidoPGS, a novel approach based on fine-mapping principles.
• To enable PGS computation using only summary-level Genome-wide association studies (GWAS) data without requiring test data for tuning.

Main Methods:

• RápidoPGS utilizes fine-mapping principles.
• The method requires only summary-level GWAS datasets.
• Parameter tuning is performed without the need for independent test data.

Main Results:

• RápidoPGS is up to 17,000-fold faster than existing methods like LDpred2, PRScs, and SBayesR.
• The method shows slightly reduced performance compared to some existing methods for case-control datasets (median r2 differences: -0.0092, -0.0042, 0.0064).
• RápidoPGS offers significantly reduced computational requirements.

Conclusions:

• RápidoPGS provides a computationally efficient alternative for calculating polygenic scores.
• The method is suitable for large-scale genetic studies where computational resources are a constraint.
• RápidoPGS is implemented as an R package and is readily available.