CARAT-GxG: CUDA-Accelerated Regression Analysis Toolkit for Large-Scale Gene–Gene Interaction with GPU Computing System

In genome-wide association studies (GWAS), regression analysis has been most commonly used to establish an association between a phenotype and genetic variants, such as single nucleotide polymorphism (SNP). However, most applications of regression analysis have been restricted to the investigation of single marker because of the large computational burden. Thus, there have been limited applications of regression analysis to multiple SNPs, including gene–gene interaction (GGI) in large-scale GWAS data. In order to overcome this limitation, we propose CARAT-GxG, a GPU computing system-oriented toolkit, for performing regression analysis with GGI using CUDA (compute unified device architecture). Compared to other methods, CARAT-GxG achieved almost 700-fold execution speed and delivered highly reliable results through our GPU-specific optimization techniques. In addition, it was possible to achieve almost-linear speed acceleration with the application of a GPU computing system, which is implemented by the TORQUE Resource Manager. We expect that CARAT-GxG will enable large-scale regression analysis with GGI for GWAS data.

Download Full-text

High-performance epistasis detection in quantitative trait GWAS

The International Journal of High Performance Computing Applications ◽

10.1177/1094342016658110 ◽

2016 ◽

Vol 32 (3) ◽

pp. 321-336 ◽

Cited By ~ 6

Author(s):

Nathan T Weeks ◽

Glenn R Luecke ◽

Brandon M Groth ◽

Marina Kraeva ◽

Li Ma ◽

...

Keyword(s):

Load Balancing ◽

Quantitative Trait ◽

Large Scale ◽

Association Studies ◽

Large Data ◽

Large Data Sets ◽

Gwas Data ◽

Data Sets ◽

Genome Wide Association Studies ◽

Data Set

epiSNP is a program for identifying pairwise single nucleotide polymorphism (SNP) interactions (epistasis) in quantitative-trait genome-wide association studies (GWAS). A parallel MPI version (EPISNPmpi) was created in 2008 to address this computationally expensive analysis on large data sets with many quantitative traits and SNP markers. However, the falling cost of genotyping has led to an explosion of large-scale GWAS data sets that challenge EPISNPmpi’s ability to compute results in a reasonable amount of time. Therefore, we optimized epiSNP for modern multi-core and highly parallel many-core processors to efficiently handle these large data sets. This paper describes the serial optimizations, dynamic load balancing using MPI-3 RMA operations, and shared-memory parallelization with OpenMP to further enhance load balancing and allow execution on the Intel Xeon Phi coprocessor (MIC). For a large GWAS data set, our optimizations provided a 38.43× speedup over EPISNPmpi on 126 nodes using 2 MICs on TACC’s Stampede Supercomputer. We also describe a Coarray Fortran (CAF) version that demonstrates the suitability of PGAS languages for problems with this computational pattern. We show that the Coarray version performs competitively with the MPI version on the NERSC Edison Cray XC30 supercomputer. Finally, the performance benefits of hyper-threading for this application on Edison (average 1.35× speedup) are demonstrated.

Download Full-text

Optimized permutation testing for information theoretic measures of multi-gene interactions

BMC Bioinformatics ◽

10.1186/s12859-021-04107-6 ◽

2021 ◽

Vol 22 (1) ◽

Author(s):

James M. Kunert-Graf ◽

Nikita A. Sakhanenko ◽

David J. Galas

Keyword(s):

Large Scale ◽

Permutation Test ◽

Association Studies ◽

Genome Wide Association Studies ◽

Permutation Testing ◽

Exact Test ◽

Information Theoretic ◽

Information Theoretic Measures ◽

Full Analysis ◽

Computational Bottleneck

Abstract Background Permutation testing is often considered the “gold standard” for multi-test significance analysis, as it is an exact test requiring few assumptions about the distribution being computed. However, it can be computationally very expensive, particularly in its naive form in which the full analysis pipeline is re-run after permuting the phenotype labels. This can become intractable in multi-locus genome-wide association studies (GWAS), in which the number of potential interactions to be tested is combinatorially large. Results In this paper, we develop an approach for permutation testing in multi-locus GWAS, specifically focusing on SNP–SNP-phenotype interactions using multivariable measures that can be computed from frequency count tables, such as those based in Information Theory. We find that the computational bottleneck in this process is the construction of the count tables themselves, and that this step can be eliminated at each iteration of the permutation testing by transforming the count tables directly. This leads to a speed-up by a factor of over 103 for a typical permutation test compared to the naive approach. Additionally, this approach is insensitive to the number of samples making it suitable for datasets with large number of samples. Conclusions The proliferation of large-scale datasets with genotype data for hundreds of thousands of individuals enables new and more powerful approaches for the detection of multi-locus genotype-phenotype interactions. Our approach significantly improves the computational tractability of permutation testing for these studies. Moreover, our approach is insensitive to the large number of samples in these modern datasets. The code for performing these computations and replicating the figures in this paper is freely available at https://github.com/kunert/permute-counts.

Download Full-text

Power comparison of Cochran-Armitage trend test against allelic and genotypic tests in large-scale case-control genetic association studies

Statistical Methods in Medical Research ◽

10.1177/0962280216683979 ◽

2016 ◽

Vol 27 (9) ◽

pp. 2657-2673 ◽

Cited By ~ 2

Author(s):

Mathieu Emily

Keyword(s):

Large Scale ◽

Association Studies ◽

Disease Model ◽

Trend Test ◽

Genome Wide Association Studies ◽

Power Functions ◽

Power Comparison ◽

Powerful Test ◽

Armitage Trend Test ◽

Mode Of Inheritance

The Cochran-Armitage trend test (CA) has become a standard procedure for association testing in large-scale genome-wide association studies (GWAS). However, when the disease model is unknown, there is no consensus on the most powerful test to be used between CA, allelic, and genotypic tests. In this article, we tackle the question of whether CA is best suited to single-locus scanning in GWAS and propose a power comparison of CA against allelic and genotypic tests. Our approach relies on the evaluation of the Taylor decompositions of non-centrality parameters, thus allowing an analytical comparison of the power functions of the tests. Compared to simulation-based comparison, our approach offers the advantage of simultaneously accounting for the multidimensionality of the set of features involved in power functions. Although power for CA depends on the sample size, the case-to-control ratio and the minor allelic frequency (MAF), our results first show that it is largely influenced by the mode of inheritance and a deviation from Hardy–Weinberg Equilibrium (HWE). Furthermore, when compared to other tests, CA is shown to be the most powerful test under a multiplicative disease model or when the single-nucleotide polymorphism largely deviates from HWE. In all other situations, CA lacks in power and differences can be substantial, especially for the recessive mode of inheritance. Finally, our results are illustrated by the comparison of the performances of the statistics in two genome scans.

Download Full-text

Comorbidities and Susceptibility to COVID-19: A Generalized Gene Set Data Mining Approach

Journal of Clinical Medicine ◽

10.3390/jcm10081666 ◽

2021 ◽

Vol 10 (8) ◽

pp. 1666

Author(s):

Micaela F. Beckman ◽

Farah Bahrani Mougeot ◽

Jean-Luc C. Mougeot

Keyword(s):

Protein Interactions ◽

Association Studies ◽

Meta Analysis ◽

Holistic Approach ◽

Gene Interaction ◽

Snp Analysis ◽

Genome Wide Association Studies ◽

Nucleotide Polymorphisms ◽

Protein Protein Interactions ◽

Data Mining Approach

The COVID-19 pandemic has led to over 2.26 million deaths for almost 104 million confirmed cases worldwide, as of 4 February 2021 (WHO). Risk factors include pre-existing conditions such as cancer, cardiovascular disease, diabetes, and obesity. Although several vaccines have been deployed, there are few alternative anti-viral treatments available in the case of reduced or non-existent vaccine protection. Adopting a long-term holistic approach to cope with the COVID-19 pandemic appears critical with the emergence of novel and more infectious SARS-CoV-2 variants. Our objective was to identify comorbidity-associated single nucleotide polymorphisms (SNPs), potentially conferring increased susceptibility to SARS-CoV-2 infection using a computational meta-analysis approach. SNP datasets were downloaded from a publicly available genome-wide association studies (GWAS) catalog for 141 of 258 candidate COVID-19 comorbidities. Gene-level SNP analysis was performed to identify significant pathways by using the program MAGMA. An SNP annotation program was used to analyze MAGMA-identified genes. Differential gene expression was determined for significant genes across 30 general tissue types using the Functional and Annotation Mapping of GWAS online tool GENE2FUNC. COVID-19 comorbidities (n = 22) from six disease categories were found to have significant associated pathways, validated by Q–Q plots (p < 0.05). Protein–protein interactions of significant (p < 0.05) differentially expressed genes were visualized with the STRING program. Gene interaction networks were found to be relevant to SARS and influenza pathogenesis. In conclusion, we were able to identify the pathways potentially affected by or affecting SARS-CoV-2 infection in underlying medical conditions likely to confer susceptibility and/or the severity of COVID-19. Our findings have implications in future COVID-19 experimental research and treatment development.

Download Full-text

GWASpro: a high-performance genome-wide association analysis server

Bioinformatics ◽

10.1093/bioinformatics/bty989 ◽

2018 ◽

Vol 35 (14) ◽

pp. 2512-2514 ◽

Cited By ~ 4

Author(s):

Bongsong Kim ◽

Xinbin Dai ◽

Wenchao Zhang ◽

Zhaohong Zhuang ◽

Darlene L Sanchez ◽

...

Keyword(s):

High Performance ◽

Large Scale ◽

Linear Mixed Model ◽

Association Studies ◽

Learning Curves ◽

Experimental Designs ◽

Genome Wide Association ◽

Supplementary Information ◽

Genome Wide Association Studies ◽

Genome Wide

Abstract Summary We present GWASpro, a high-performance web server for the analyses of large-scale genome-wide association studies (GWAS). GWASpro was developed to provide data analyses for large-scale molecular genetic data, coupled with complex replicated experimental designs such as found in plant science investigations and to overcome the steep learning curves of existing GWAS software tools. GWASpro supports building complex design matrices, by which complex experimental designs that may include replications, treatments, locations and times, can be accounted for in the linear mixed model. GWASpro is optimized to handle GWAS data that may consist of up to 10 million markers and 10 000 samples from replicable lines or hybrids. GWASpro provides an interface that significantly reduces the learning curve for new GWAS investigators. Availability and implementation GWASpro is freely available at https://bioinfo.noble.org/GWASPRO. Supplementary information Supplementary data are available at Bioinformatics online.

Download Full-text

A description of large-scale metabolomics studies: increasing value by combining metabolomics with genome-wide SNP genotyping and transcriptional profiling

Journal of Endocrinology ◽

10.1530/joe-12-0144 ◽

2012 ◽

Vol 215 (1) ◽

pp. 17-28 ◽

Cited By ~ 18

Author(s):

Georg Homuth ◽

Alexander Teumer ◽

Uwe Völker ◽

Matthias Nauck

Keyword(s):

Blood Cells ◽

Large Scale ◽

Genetic Factors ◽

Association Studies ◽

Transcriptional Profiling ◽

Genome Wide Association Studies ◽

Protein Levels ◽

Future Developments ◽

Genome Wide ◽

Metabolome Data

The metabolome, defined as the reflection of metabolic dynamics derived from parameters measured primarily in easily accessible body fluids such as serum, plasma, and urine, can be considered as the omics data pool that is closest to the phenotype because it integrates genetic influences as well as nongenetic factors. Metabolic traits can be related to genetic polymorphisms in genome-wide association studies, enabling the identification of underlying genetic factors, as well as to specific phenotypes, resulting in the identification of metabolome signatures primarily caused by nongenetic factors. Similarly, correlation of metabolome data with transcriptional or/and proteome profiles of blood cells also produces valuable data, by revealing associations between metabolic changes and mRNA and protein levels. In the last years, the progress in correlating genetic variation and metabolome profiles was most impressive. This review will therefore try to summarize the most important of these studies and give an outlook on future developments.

Download Full-text

A practical approach to adjusting for population stratification in genome-wide association studies: principal components and propensity scores (PCAPS)

Statistical Applications in Genetics and Molecular Biology ◽

10.1515/sagmb-2017-0054 ◽

2018 ◽

Vol 17 (6) ◽

Cited By ~ 2

Author(s):

Huaqing Zhao ◽

Nandita Mitra ◽

Peter A. Kanetsky ◽

Katherine L. Nathanson ◽

Timothy R. Rebbeck

Keyword(s):

Principal Components ◽

Population Stratification ◽

Propensity Scores ◽

Association Studies ◽

Germ Cell Tumors ◽

Gwas Data ◽

Genome Wide Association ◽

Genome Wide Association Studies ◽

Testicular Germ Cell ◽

Genome Wide

Abstract Genome-wide association studies (GWAS) are susceptible to bias due to population stratification (PS). The most widely used method to correct bias due to PS is principal components (PCs) analysis (PCA), but there is no objective method to guide which PCs to include as covariates. Often, the ten PCs with the highest eigenvalues are included to adjust for PS. This selection is arbitrary, and patterns of local linkage disequilibrium may affect PCA corrections. To address these limitations, we estimate genomic propensity scores based on all statistically significant PCs selected by the Tracy-Widom (TW) statistic. We compare a principal components and propensity scores (PCAPS) approach to PCA and EMMAX using simulated GWAS data under no, moderate, and severe PS. PCAPS reduced spurious genetic associations regardless of the degree of PS, resulting in odds ratio (OR) estimates closer to the true OR. We illustrate our PCAPS method using GWAS data from a study of testicular germ cell tumors. PCAPS provided a more conservative adjustment than PCA. Advantages of the PCAPS approach include reduction of bias compared to PCA, consistent selection of propensity scores to adjust for PS, the potential ability to handle outliers, and ease of implementation using existing software packages.

Download Full-text

Better estimation of SNP heritability from summary statistics provides a new understanding of the genetic architecture of complex traits

10.1101/284976 ◽

2018 ◽

Cited By ~ 6

Author(s):

Doug Speed ◽

David J Balding

Keyword(s):

Complex Traits ◽

Genetic Architecture ◽

Large Scale ◽

Association Studies ◽

Genome Wide Association Studies ◽

Summary Statistics ◽

Confounding Bias ◽

Conserved Regions ◽

Genome Wide ◽

Variation Explained

LD Score Regression (LDSC) has been widely applied to the results of genome-wide association studies. However, its estimates of SNP heritability are derived from an unrealistic model in which each SNP is expected to contribute equal heritability. As a consequence, LDSC tends to over-estimate confounding bias, under-estimate the total phenotypic variation explained by SNPs, and provide misleading estimates of the heritability enrichment of SNP categories. Therefore, we present SumHer, software for estimating SNP heritability from summary statistics using more realistic heritability models. After demonstrating its superiority over LDSC, we apply SumHer to the results of 24 large-scale association studies (average sample size 121 000). First we show that these studies have tended to substantially over-correct for confounding, and as a result the number of genome-wide significant loci has under-reported by about 20%. Next we estimate enrichment for 24 categories of SNPs defined by functional annotations. A previous study using LDSC reported that conserved regions were 13-fold enriched, and found a further twelve categories with above 2-fold enrichment. By contrast, our analysis using SumHer finds that conserved regions are only 1.6-fold (SD 0.06) enriched, and that no category has enrichment above 1.7-fold. SumHer provides an improved understanding of the genetic architecture of complex traits, which enables more efficient analysis of future genetic data.

Download Full-text

RAFFI: Accurate and fast familial relationship inference in large scale biobank studies using RaPID

PLoS Genetics ◽

10.1371/journal.pgen.1009315 ◽

2021 ◽

Vol 17 (1) ◽

pp. e1009315

Author(s):

Ardalan Naseri ◽

Junjie Shi ◽

Xihong Lin ◽

Shaojie Zhang ◽

Degui Zhi

Keyword(s):

Large Scale ◽

Association Studies ◽

Scale Up ◽

Data Driven ◽

Genome Wide Association Studies ◽

Inference Method ◽

Genome Wide ◽

Familial Relationship ◽

Kinship Coefficients ◽

Data Driven Approach

Inference of relationships from whole-genome genetic data of a cohort is a crucial prerequisite for genome-wide association studies. Typically, relationships are inferred by computing the kinship coefficients (ϕ) and the genome-wide probability of zero IBD sharing (π0) among all pairs of individuals. Current leading methods are based on pairwise comparisons, which may not scale up to very large cohorts (e.g., sample size >1 million). Here, we propose an efficient relationship inference method, RAFFI. RAFFI leverages the efficient RaPID method to call IBD segments first, then estimate the ϕ and π0 from detected IBD segments. This inference is achieved by a data-driven approach that adjusts the estimation based on phasing quality and genotyping quality. Using simulations, we showed that RAFFI is robust against phasing/genotyping errors, admix events, and varying marker densities, and achieves higher accuracy compared to KING, the current leading method, especially for more distant relatives. When applied to the phased UK Biobank data with ~500K individuals, RAFFI is approximately 18 times faster than KING. We expect RAFFI will offer fast and accurate relatedness inference for even larger cohorts.

Download Full-text

297 GWAS for complex models accounting for populations structure with GBLUP and ssGBLUP

Journal of Animal Science ◽

10.1093/jas/skaa278.057 ◽

2020 ◽

Vol 98 (Supplement_4) ◽

pp. 32-32

Author(s):

Juan P Steibel ◽

Ignacio Aguilar

Keyword(s):

Hypothesis Testing ◽

Large Scale ◽

Mixed Model ◽

Prediction Models ◽

Association Studies ◽

Least Square ◽

Type I ◽

Phenotypic Variance ◽

Genome Wide Association Studies ◽

Formal Hypothesis Testing

Abstract Genomic Best Linear Unbiased Prediction (GBLUP) is the method of choice for incorporating genomic information into the genetic evaluation of livestock species. Furthermore, single step GBLUP (ssGBLUP) is adopted by many breeders’ associations and private entities managing large scale breeding programs. While prediction of breeding values remains the primary use of genomic markers in animal breeding, a secondary interest focuses on performing genome-wide association studies (GWAS). The goal of GWAS is to uncover genomic regions that harbor variants that explain a large proportion of the phenotypic variance, and thus become candidates for discovering and studying causative variants. Several methods have been proposed and successfully applied for embedding GWAS into genomic prediction models. Most methods commonly avoid formal hypothesis testing and resort to estimation of SNP effects, relying on visual inspection of graphical outputs to determine candidate regions. However, with the advent of high throughput phenomics and transcriptomics, a more formal testing approach with automatic discovery thresholds is more appealing. In this work we present the methodological details of a method for performing formal hypothesis testing for GWAS in GBLUP models. First, we present the method and its equivalencies and differences with other GWAS methods. Moreover, we demonstrate through simulation analyses that the proposed method controls type I error rate at the nominal level. Second, we demonstrate two possible computational implementations based on mixed model equations for ssGBLUP and based on the generalized least square equations (GLS). We show that ssGBLUP can deal with datasets with extremely large number of animals and markers and with multiple traits. GLS implementations are well suited for dealing with smaller number of animals with tens of thousands of phenotypes. Third, we show several useful extensions, such as: testing multiple markers at once, testing pleiotropic effects and testing association of social genetic effects.

Download Full-text