Solving Sequences of Generalized Least-Squares Problems on Multi-threaded Architectures

TL;DR

This paper introduces high-performance algorithms for solving large sequences of generalized least-squares problems in genome-wide association studies, leveraging multi-core architectures to significantly accelerate computations.

Contribution

The paper presents novel shared-memory algorithms that exploit domain knowledge and parallelism, achieving up to 50-fold speedups over existing libraries like GenABEL.

Findings

50-fold speedup compared to GenABEL

Enables genome studies of unprecedented size

Efficient handling of terabytes of data

Abstract

Generalized linear mixed-effects models in the context of genome-wide association studies (GWAS) represent a formidable computational challenge: the solution of millions of correlated generalized least-squares problems, and the processing of terabytes of data. We present high performance in-core and out-of-core shared-memory algorithms for GWAS: By taking advantage of domain-specific knowledge, exploiting multi-core parallelism, and handling data efficiently, our algorithms attain unequalled performance. When compared to GenABEL, one of the most widely used libraries for GWAS, on a 12-core processor we obtain 50-fold speedups. As a consequence, our routines enable genome studies of unprecedented size.