High-performance epistasis detection in quantitative trait GWAS

Weeks, Nathan T.; Luecke, Glenn R.; Groth, Brandon M.; Kraeva, Marina; Ma, Li; Kramer, Luke M.; Koltes, James E.; Reecy, James M.

doi:10.1177/1094342016658110

Cited by 25 publications

(17 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The AGLS method combines the least squares (LS) tests implemented by EPISNP1mpi (Ma et al, 2008; Weeks et al, 2016) with the estimated breeding values from routine genetic evaluation using the entire U.S. Holstein population. The statistical model was…”

Section: Methodsmentioning

confidence: 99%

“…In addition to being a computationally efficient method for sample stratification correction for large samples, the AGLS method implemented by EPSNPmpi (Ma et al, 2008; Weeks et al, 2016) offers tests and estimates unavailable from BOLT-LMM, including dominance test, estimates of allelic and dominance effects, and estimates of allele and genotypic frequencies of each SNP. Additive effects of each SNP were estimated using three measures, the average effect of gene substitution, allelic mean, and allelic effect of each allele based on quantitative genetics definitions (Falconer and Mackay, 1996; Mao et al, 2007; Da, 2015).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A Large-Scale Genome-Wide Association Study in U.S. Holstein Cattle

et al. 2019

Self Cite

View full text Add to dashboard Cite

Genome-wide association study (GWAS) is a powerful approach to identify genomic regions and genetic variants associated with phenotypes. However, only limited mutual confirmation from different studies is available. We conducted a large-scale GWAS using 294,079 first-lactation Holstein cows and identified new additive and dominance effects on five production traits, three fertility traits, and somatic cell score. Four chromosomes had the most significant SNP effects on the five production traits, a Chr14 region containing DGAT1 mostly had positive effects on fat yield and negative effects on milk and protein yields, the 88.07–89.60 Mb region of Chr06 with SLC4A4, GC, NPFFR2 , and ADAMTS3 for milk and protein yields, the 30.03–36.67 Mb region of Chr20 with C6 and GHR for milk yield, and the 88.19–88.88 Mb region with ABCC9 as well as the 91.13–94.62 Mb region of Chr05 with PLEKHA5, MGST1, SLC15A5 , and EPS8 for fat yield. For fertility traits, the SNP in GC of Chr06, and the SNPs in the 65.02–69.43 Mb region of Chr01 with COX17, ILDR1 , and KALRN had the most significant effects for daughter pregnancy rate and cow conception rate, whereas SNPs in AFF1 of Chr06, the 47.54–52.79 Mb region of Chr07, TSPAN4 of Chr29, and NPAS1 of Chr18 had the most significant effects for heifer conception rate. For somatic cell score, GC of Chr06 and PRLR of Chr20 had the most significant effects. A small number of dominance effects were detected for the production traits with far lower statistical significance than the additive effects and for fertility traits with similar statistical significance as the additive effects. Analysis of allelic effects revealed the presence of uni-allelic, asymmetric, and symmetric SNP effects and found the previously reported DGAT1 antagonism was an extreme antagonistic pleiotropy between fat yield and milk and protein yields among all SNPs in this study.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

A Large-Scale Genome-Wide Association Study in U.S. Holstein Cattle

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…[20] performs 2-way analyses on up to 126 nodes of the Intel Phi-based Stampede system (cf. [21]). [22] considers 2-way computations on thousands of compute cores with good scalability and good absolute performance on conventional CPUs.…”

Section: Introductionmentioning

confidence: 99%

Parallel accelerated vector similarity calculations for genomics applications

et al. 2018

View full text Add to dashboard Cite

The surge in availability of genomic data holds promise for enabling determination of genetic causes of observed individual traits, with applications to problems such as discovery of the genetic roots of phenotypes, be they molecular phenotypes such as gene expression or metabolite concentrations, or complex phenotypes such as diseases. However, the growing sizes of these datasets and the quadratic, cubic or higher scaling characteristics of the relevant algorithms pose a serious computational challenge necessitating use of leadership scale computing. In this paper we describe a new approach to performing vector similarity metrics calculations, suitable for parallel systems equipped with graphics processing units (GPUs) or Intel Xeon Phi processors. Our primary focus is the Proportional Similarity metric applied to Genome Wide Association Studies (GWAS) and Phenome Wide Association Studies (PheWAS). We describe the implementation of the algorithms on accelerated processors, methods used for eliminating redundant calculations due to symmetries, and techniques for efficient mapping of the calculations to many-node parallel systems. Results are presented demonstrating high per-node performance and parallel scalability with rates of more than five quadrillion (5 × 10 15 ) elementwise comparisons achieved per second on the ORNL Titan system. In a companion paper we describe corresponding techniques applied to calculations of the Custom Correlation Coefficient for comparative genomics applications.

show abstract

“…The evaluation was performed with a bioinformatics Fortran application comprised of 2000+ lines of code, called epiSNP, written by Ma (Ma et al 2008) and optimized by Weeks (Weeks et al 2016) aided by Allinea MAP and Intel's PFLET. There are several versions of the optimized epiSNP, including a serial version and a hybrid MPI+OpenMP version.…”

Section: Introductionmentioning

confidence: 99%

“…All runs were performed on Iowa State University's CyEnce Cluster (see http://www.hpc.iastate.edu/systems). The MPI+OpenMP epiSNP ran with this configuration had an approximate runtime of 7 hours with the 4.4GB data set used in Weeks (Weeks et al 2016).…”

Section: Introductionmentioning

confidence: 99%

Comparing Allinea"s and Intel"s Performance Tools for HPC

2017

25th High Performance Computing Symposium (HPC 2017)

View full text Add to dashboard Cite

To efficiently use HPC machines, it is critical to optimize applications for high performance. To accomplish this, HPC developers must utilize performance tools to find and correct performance problems within large, complex scientific applications. Allinea and Intel offer vendor-supported performance tools that are regularly updated to capture important performance metrics on the latest hardware. In this paper, the authors evaluated and compared Allinea's MAP performance tool and Intel's performance tools to aid in application optimization. The authors found that Allinea's MAP provided useful performance metrics necessary to diagnose and fix performance problems using an intuitive, easy-to-use user interface. Intel's performance tools provided a more detailed and customizable view of application performance, at the expense of a more complicated user interface. The comparison presented in this paper will help HPC developers decide which performance tool is best for them.

show abstract

High-performance epistasis detection in quantitative trait GWAS

Cited by 25 publications

References 13 publications

A Large-Scale Genome-Wide Association Study in U.S. Holstein Cattle

A Large-Scale Genome-Wide Association Study in U.S. Holstein Cattle

Parallel accelerated vector similarity calculations for genomics applications

Comparing Allinea"s and Intel"s Performance Tools for HPC

Contact Info

Product

Resources

About