Phenotypes and Genotypes by Florian Frommlet Małgorzata Bogdan & David Ramsey

Author:Florian Frommlet, Małgorzata Bogdan & David Ramsey
Language: eng
Format: epub
Publisher: Springer London, London

In 2005, even before any GWAS had been carried out, Hirschhorn and Daly [57] gave a concise summary of the problems involved in them. They gave an argument for why GWAS are useful and discussed the inherent multiple testing problem in depth, along with strategies of dealing with it. Several other reviews dealing with the statistical aspects involved in GWAS are available, see for example [20, 79, 134].

The first task of GWAS is to infer the genotype of SNPs from the DNA array image data. A brief and non-exhaustive survey of methods for inferring genotypes (genotype calling) is given in Sect. 5.2.1. Often, genotype calling for a specific SNP is not successful for all individuals, but one would still like to use such SNPs in an analysis. Also, one might be interested in the genotypes at sites which have not been previously genotyped, and sometimes one is actually interested in inferring not only genotypes, but haplotypes. Strategies for imputing genotypes and haplotype phasing are discussed in Sect. 5.2.2.

After genotypes have been called, the actual association analysis is performed. In Sect. 5.3, the most common statistical approaches based on single markers are reviewed. In terms of a frequentist approach to the analysis of GWAS, it has become standard to start by computing p-values from single marker tests and then consider markers associated with realizations of a test statistic above some threshold reflecting correction for multiple testing [40]. However, as pointed out for example in [57], genome-wide association studies are based on the assumption that common diseases are associated with common variants. This hypothesis suggests that common diseases result from a moderate number of small common genetic effects. Given this hypothesis is true, such single marker tests are bound to fail, as was explained exhaustively in [48]. Recently, a number of approaches have been introduced to analyze GWAS using strategies for model selection, which will be presented in Sect. 5.4.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.