Unlocking Personalized Healthcare on Modern CPUs/GPUs: Three-way Gene Interaction Study

TL;DR

This paper introduces optimized methods for detecting three-way gene interactions on modern CPUs and GPUs, leveraging architectural insights to significantly improve performance in personalized healthcare genomics applications.

Contribution

It presents novel GPU and CPU algorithms for epistasis detection, validated by extensive architectural analysis and outperforming existing methods in speed.

Findings

Achieved an average speedup of 3.9× over state-of-the-art methods.

Surpassed performance on 13 CPU and GPU devices from main vendors.

Maximum speedup of 10.6× on Intel UHD P630 GPU.

Abstract

Developments in Genome-Wide Association Studies have led to the increasing notion that future healthcare techniques will be personalized to the patient, by relying on genetic tests to determine the risk of developing a disease. To this end, the detection of gene interactions that cause complex diseases constitutes an important application. Similarly to many applications in this field, extensive data sets containing genetic information for a series of patients are used (such as Single-Nucleotide Polymorphisms), leading to high computational complexity and memory utilization, thus constituting a major challenge when targeting high-performance execution in modern computing systems. To close this gap, this work proposes several novel approaches for the detection of three-way gene interactions in modern CPUs and GPUs, making use of different optimizations to fully exploit the target architectures. Crucial insights from the Cache-Aware Roofline Model are used to ensure the suitability of the applications to the computing devices. An extensive study of the architectural features of 13 CPU and GPU devices from all main vendors is also presented, allowing to understand the features relevant to obtain high-performance in this bioinformatics domain. To the best of our knowledge, this study is the first to perform such evaluation for epistasis detection. The proposed approaches are able to surpass the performance of state-of-the-art works in the tested platforms, achieving an average speedup of 3.9$\times$ (7.3$\times$ on CPUs and 2.8$\times$ on GPUs) and maximum speedup of 10.6$\times$ on Intel UHD P630 GPU.