ABSTRACT
The improved accessibility of gene sequencing technologies has led to creation of huge datasets, i.e. patient records related to certain human diseases (phenotypes). Hence, deriving fast and accurate algorithms for efficiently processing these datasets is a paramount concern to enable some key healthcare scenarios, such as personalizing treatments, explaining the occurrence of and/or susceptibility to complex conditions and reducing the spread of infectious diseases. This is especially true for high-order epistasis detection, one of the most computationally challenging problems in bioinformatics, where associations between a given phenotype and single nucleotide polymorphisms (SNPs) of a population can often only be uncovered through evaluation of a large number of SNP combinations. To tackle this challenge, we propose a novel fourth-order epistasis detection algorithm that leverages tensor processing capabilities of two distinct accelerator architectures by efficiently mapping core computations related to processing quads of SNPs to binary tensor-accelerated matrix operations. Experimental results show that the proposed approach delivers very high performance even in single-GPU environments, e.g., 27.8 and 90.9 tera quads of SNPs per second, scaled to the sample size, were processed on Titan RTX (Turing) and A100 (Ampere) PCIe GPUs, respectively. Being the first approach that exploits tensor cores for accelerating searches with interaction order above three, the proposed method achieved a performance of up to 835.4 tera quads of SNPs per second on the 8-GPU HGX A100 server, which represents performance two or more orders of magnitude higher than that of related art.
- George E. Andrews, Richard Askey, and Ranjan Roy. 1999. The Gamma and Beta Functions. Cambridge University Press, 1–60.Google Scholar
- Arash Bayat and others. 2021. Fast and accurate exhaustive higher-order epistasis search with BitEpi. Scientific Reports 11, 1 (05 Aug 2021), 15923.Google Scholar
- Rafael Campos and others. 2020. Heterogeneous CPU+iGPU Processing for Efficient Epistasis Detection. In Euro-Par 2020: Parallel Processing. Springer International Publishing, Cham, 613–628.Google Scholar
- Gregory F. Cooper and Edward Herskovits. 1992. A Bayesian method for the induction of probabilistic networks from data. Machine Learning 9, 4 (01 Oct 1992), 309–347.Google Scholar
- Jorge González-Domínguez, Sabela Ramos, Juan Touriño, and Bertil Schmidt. 2016. Parallel Pairwise Epistasis Detection on Heterogeneous Computing Architectures. IEEE Transactions on Parallel and Distributed Systems 27, 8 (Aug 2016), 2329–2340.Google ScholarDigital Library
- Jorge González-Domínguez and Bertil Schmidt. 2015. GPU-accelerated exhaustive search for third-order epistatic interactions in case–control studies. Journal of Computational Science 8 (2015), 93 – 100.Google ScholarCross Ref
- Jorge González-Domínguez, Jan Christian Kässens, Lars Wienbrandt, and Bertil Schmidt. 2015. Large-scale genome-wide association studies on a GPU cluster using a CUDA-accelerated PGAS programming model. The International Journal of High Performance Computing Applications 29, 4(2015), 506–510.Google ScholarDigital Library
- Fernando Pires Hartwig and others. 2014. Evidence for an Epistatic Effect between TP53 R72P and MDM2 T309G SNPs in HIV Infection: A Cross-Sectional Study in Women from South Brazil. PLOS ONE 9, 2 (02 2014), 1–11.Google Scholar
- Cindy Im and others. 2018. Genome-wide search for higher order epistasis as modifiers of treatment effects on bone mineral density in childhood cancer survivors. European journal of human genetics : EJHG 26, 2 (2018), 275–286.Google ScholarCross Ref
- Wayne Joubert and others. 2018. Attacking the Opioid Epidemic: Determining the Epistatic and Pleiotropic Genetic Architectures for Chronic Pain and Opioid Addiction. In Proceedings of the International Conference for High Performance Computing, Networking, Storage, and Analysis(SC ’18). IEEE Press, Piscataway, NJ, USA, Article 57, 14 pages.Google Scholar
- Daniel Jünger, Christian Hundt, Jorge González Domínguez, and Bertil Schmidt. 2017. Speed and accuracy improvement of higher-order epistasis detection on CUDA-enabled GPUs. Cluster Computing 20, 3 (01 Sep 2017), 1899–1908.Google Scholar
- Jan C. Kässens, Jorge González-Domínguez, Lars Wienbrandt, and Bertil Schmidt. 2014. UPC++ for bioinformatics: A case study using genome-wide association studies. In 2014 IEEE International Conference on Cluster Computing (CLUSTER). 248–256.Google ScholarCross Ref
- Trudy FC Mackay and Jason H Moore. 2014. Why epistasis is important for tackling complex human disease genetics. Genome medicine (6):42(2014), 1–3.Google Scholar
- Ricardo Nobre, Aleksandar Ilic, Sergio Santander-Jiménez, and Leonel Sousa. 2020. Exploring the Binary Precision Capabilities of Tensor Cores for Epistasis Detection. In 2020 International Parallel and Distributed Processing Symposium (IPDPS 2020).Google Scholar
- Ricardo Nobre, Aleksandar Ilic, Sergio Santander-Jiménez, and Leonel Sousa. 2021. Fourth-Order Exhaustive Epistasis Detection for the XPU Era. In 50th International Conference on Parallel Processing(ICPP 2021). Association for Computing Machinery, New York, NY, USA, Article 27, 10 pages.Google Scholar
- Ricardo Nobre, Aleksandar Ilic, Sergio Santander-Jiménez, and Leonel Sousa. 2021. Retargeting Tensor Accelerators for Epistasis Detection. IEEE Transactions on Parallel and Distributed Systems 32, 9 (2021), 2160–2174.Google ScholarCross Ref
- NVIDIA Corporation. NVIDIA A100 Tensor Core GPU Architecture. https://www.nvidia.com/content/dam/en-zz/Solutions/Data-Center/nvidia-ampere-architecture-whitepaper.pdf. [Online; visited January-2022].Google Scholar
- [18] NVIDIA Corporation.https://www.nvidia.com/en-us/data-center/hgx[Online; visited January-2022].Google Scholar
- NVIDIA Corporation. NVIDIA Turing GPU Architecture Whitepaper. https://www.nvidia.com/content/dam/en-zz/Solutions/design-visualization/technologies/turing-architecture/NVIDIA-Turing-Architecture-Whitepaper.pdf. [Online; visited January-2022].Google Scholar
- Christian Ponte-Fernández, Jorge González-Domínguez, and María J Martín. 2020. Fast search of third-order epistatic interactions on CPU and GPU clusters. The International Journal of High Performance Computing Applications 34, 1(2020), 20–29. http://dx.doi.org/10.1177/1094342019852128Google ScholarDigital Library
- Gaspar Ribeiro, Nuno Neves, Sergio Santander-Jiménez, and Aleksandar Ilic. 2021. HEDAcc: FPGA-based Accelerator for High-order Epistasis Detection. In 2021 IEEE 29th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM). 124–132.Google Scholar
- Jiya Sun and others. 2014. Hidden Risk Genes with High-Order Intragenic Epistasis in Alzheimer’s Disease. Journal of Alzheimer’s disease : JAD 41 (04 2014).Google Scholar
- Yingxia Sun and others. 2017. epiACO - a method for identifying epistasis based on ant Colony optimization algorithm. BioData mining 10:23 (06 Jul 2017), 1–17.Google Scholar
- Xiang Wan and others. 2010. BOOST: A fast approach to detecting gene-gene interactions in genome-wide case-control studies. American journal of human genetics 87, 3 (10 Sep 2010), 325–340.Google Scholar
- Lars Wienbrandt, Jan Christian Kässens, Matthias Hübenthal, and David Ellinghaus. 2019. 1000 × faster than PLINK: Combined FPGA and GPU accelerators for logistic regression-based detection of epistasis. Journal of Computational Science 30 (2019), 183 – 193.Google ScholarCross Ref
- Ling Sing Yung, Can Yang, Xiang Wan, and Weichuan Yu. 2011. GBOOST. Bioinformatics 27, 9 (May 2011), 1309–1310.Google ScholarCross Ref
- Shawna J. Zimmerman, Cameron L. Aldridge, and Sara J. Oyler-McCance. 2020. An empirical comparison of population genetic analyses using microsatellite and SNP data for a species of conservation concern. BMC Genomics 21, 1 (01 Jun 2020), 382.Google Scholar
Index Terms
- Tensor-Accelerated Fourth-Order Epistasis Detection on GPUs
Recommendations
Fourth-Order Exhaustive Epistasis Detection for the xPU Era
ICPP '21: Proceedings of the 50th International Conference on Parallel ProcessingThe investigation of highly-efficient parallel algorithms targeting modern heterogeneous systems can provide bioinformaticians new mechanisms to find relations between genetics, phenotype and environment. This paper proposes an approach for fourth-order ...
Speed and accuracy improvement of higher-order epistasis detection on CUDA-enabled GPUs
The discovery of higher-order epistatic interactions is an important task in the field of genome wide association studies which allows for the identification of complex interaction patterns between multiple genetic markers. Some existing bruteforce ...
Accelerating 3-Way Epistasis Detection with CPU+GPU Processing
Job Scheduling Strategies for Parallel ProcessingAbstractA Single Nucleotide Polymorphism (SNP) is a DNA variation occurring when a single nucleotide differs between individuals of a species. Some conditions can be explained with a single SNP. However, the combined effect of multiple SNPs, known as ...
Comments