Nilotpal Sanyal1, Min-Tzu Lo1, Karolina Kauppi2, Srdjan Djurovic3, Ole A Andreassen4, Valen E Johnson5, Chi-Hua Chen1. 1. Department of Radiology, University of California, San Diego, La Jolla, CA, USA. 2. Department of Radiation Sciences, Umeå University, Umeå, Sweden. 3. Department of Medical Genetics, NORMENT, KG Jebsen Centre, University of Bergen, Bergen, Oslo University Hospital, Oslo, Norway. 4. Division of Mental Health and Addiction, NORMENT, KG Jebsen Centre, Oslo University Hospital and Institute of Clinical Medicine, University of Oslo, Oslo, Norway. 5. Department of Statistics, Texas A&M University, College Station, TX, USA.
Abstract
Motivation: Multiple marker analysis of the genome-wide association study (GWAS) data has gained ample attention in recent years. However, because of the ultra high-dimensionality of GWAS data, such analysis is challenging. Frequently used penalized regression methods often lead to large number of false positives, whereas Bayesian methods are computationally very expensive. Motivated to ameliorate these issues simultaneously, we consider the novel approach of using non-local priors in an iterative variable selection framework. Results: We develop a variable selection method, named, iterative non-local prior based selection for GWAS, or GWASinlps, that combines, in an iterative variable selection framework, the computational efficiency of the screen-and-select approach based on some association learning and the parsimonious uncertainty quantification provided by the use of non-local priors. The hallmark of our method is the introduction of 'structured screen-and-select' strategy, that considers hierarchical screening, which is not only based on response-predictor associations, but also based on response-response associations and concatenates variable selection within that hierarchy. Extensive simulation studies with single nucleotide polymorphisms having realistic linkage disequilibrium structures demonstrate the advantages of our computationally efficient method compared to several frequentist and Bayesian variable selection methods, in terms of true positive rate, false discovery rate, mean squared error and effect size estimation error. Further, we provide empirical power analysis useful for study design. Finally, a real GWAS data application was considered with human height as phenotype. Availability and implementation: An R-package for implementing the GWASinlps method is available at https://cran.r-project.org/web/packages/GWASinlps/index.html. Supplementary information: Supplementary data are available at Bioinformatics online.
Motivation: Multiple marker analysis of the genome-wide association study (GWAS) data has gained ample attention in recent years. However, because of the ultra high-dimensionality of GWAS data, such analysis is challenging. Frequently used penalized regression methods often lead to large number of false positives, whereas Bayesian methods are computationally very expensive. Motivated to ameliorate these issues simultaneously, we consider the novel approach of using non-local priors in an iterative variable selection framework. Results: We develop a variable selection method, named, iterative non-local prior based selection for GWAS, or GWASinlps, that combines, in an iterative variable selection framework, the computational efficiency of the screen-and-select approach based on some association learning and the parsimonious uncertainty quantification provided by the use of non-local priors. The hallmark of our method is the introduction of 'structured screen-and-select' strategy, that considers hierarchical screening, which is not only based on response-predictor associations, but also based on response-response associations and concatenates variable selection within that hierarchy. Extensive simulation studies with single nucleotide polymorphisms having realistic linkage disequilibrium structures demonstrate the advantages of our computationally efficient method compared to several frequentist and Bayesian variable selection methods, in terms of true positive rate, false discovery rate, mean squared error and effect size estimation error. Further, we provide empirical power analysis useful for study design. Finally, a real GWAS data application was considered with human height as phenotype. Availability and implementation: An R-package for implementing the GWASinlps method is available at https://cran.r-project.org/web/packages/GWASinlps/index.html. Supplementary information: Supplementary data are available at Bioinformatics online.
Authors: Leonardo Bottolo; Marc Chadeau-Hyam; David I Hastie; Tanja Zeller; Benoit Liquet; Paul Newcombe; Loic Yengo; Philipp S Wild; Arne Schillert; Andreas Ziegler; Sune F Nielsen; Adam S Butterworth; Weang Kee Ho; Raphaële Castagné; Thomas Munzel; David Tregouet; Mario Falchi; François Cambien; Børge G Nordestgaard; Fredéric Fumeron; Anne Tybjærg-Hansen; Philippe Froguel; John Danesh; Enrico Petretto; Stefan Blankenberg; Laurence Tiret; Sylvia Richardson Journal: PLoS Genet Date: 2013-08-08 Impact factor: 5.917
Authors: Yunpeng Wang; Wesley K Thompson; Andrew J Schork; Dominic Holland; Chi-Hua Chen; Francesco Bettella; Rahul S Desikan; Wen Li; Aree Witoelar; Verena Zuber; Anna Devor; Markus M Nöthen; Marcella Rietschel; Qiang Chen; Thomas Werge; Sven Cichon; Daniel R Weinberger; Srdjan Djurovic; Michael O'Donovan; Peter M Visscher; Ole A Andreassen; Anders M Dale Journal: PLoS Genet Date: 2016-01-25 Impact factor: 5.917