Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

An integrated phenomic approach to multivariate allelic association

An integrated phenomic approach to multivariate allelic association The increased feasibility of genome-wide association has resulted in association becoming the primary method used to localize genetic variants that cause phenotypic variation. Much attention has been focused on the vast multiple testing problems arising from analyzing large numbers of single nucleotide polymorphisms. However, the inflation of experiment-wise type I error rates through testing numerous phenotypes has received less attention. Multivariate analyses can be used to detect both pleiotropic effects that influence a latent common factor, and monotropic effects that operate at a variable-specific levels, whilst controlling for non-independence between phenotypes. In this study, we present a maximum likelihood approach, which combines both latent and variable-specific tests and which may be used with either individual or family data. Simulation results indicate that in the presence of factor-level association, the combined multivariate (CMV) analysis approach performs well with a minimal loss of power as compared with a univariate analysis of a factor or sum score (SS). As the deviation between the pattern of allelic effects and the factor loadings increases, the power of univariate analyses of both factor and SSs decreases dramatically, whereas the power of the CMV approach is maintained. We show the utility of the approach by examining the association between dopamine receptor D2 TaqIA and the initiation of marijuana, tranquilizers and stimulants in data from the Add Health Study. Perl scripts that takes ped and dat files as input and produces Mx scripts and data for running the CMV approach can be downloaded from www.vipbg.vcu.edu/~sarahme/WriteMx . http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Human Genetics Springer Journals

An integrated phenomic approach to multivariate allelic association

Download PDF
7 pages

Loading next page...
 
/lp/springer-journals/an-integrated-phenomic-approach-to-multivariate-allelic-association-0eKgsFYKol

References (22)

Publisher
Springer Journals
Copyright
Copyright © 2010 by Macmillan Publishers Limited
Subject
Biomedicine; Biomedicine, general; Human Genetics; Bioinformatics; Gene Expression; Cytogenetics
ISSN
1018-4813
eISSN
1476-5438
DOI
10.1038/ejhg.2009.133
Publisher site
See Article on Publisher Site

Abstract

The increased feasibility of genome-wide association has resulted in association becoming the primary method used to localize genetic variants that cause phenotypic variation. Much attention has been focused on the vast multiple testing problems arising from analyzing large numbers of single nucleotide polymorphisms. However, the inflation of experiment-wise type I error rates through testing numerous phenotypes has received less attention. Multivariate analyses can be used to detect both pleiotropic effects that influence a latent common factor, and monotropic effects that operate at a variable-specific levels, whilst controlling for non-independence between phenotypes. In this study, we present a maximum likelihood approach, which combines both latent and variable-specific tests and which may be used with either individual or family data. Simulation results indicate that in the presence of factor-level association, the combined multivariate (CMV) analysis approach performs well with a minimal loss of power as compared with a univariate analysis of a factor or sum score (SS). As the deviation between the pattern of allelic effects and the factor loadings increases, the power of univariate analyses of both factor and SSs decreases dramatically, whereas the power of the CMV approach is maintained. We show the utility of the approach by examining the association between dopamine receptor D2 TaqIA and the initiation of marijuana, tranquilizers and stimulants in data from the Add Health Study. Perl scripts that takes ped and dat files as input and produces Mx scripts and data for running the CMV approach can be downloaded from www.vipbg.vcu.edu/~sarahme/WriteMx .

Journal

European Journal of Human Genetics – Springer Journals

Published: Aug 26, 2009

There are no references for this article.