Statistical methods for detecting genotype-phenotype association in the presence of environmental covariates
This thesis shows how statistical methods based on logistic regression models can
be used to analyze and interpret biological data. In genome-wide association stud-
ies, the aim is to detect association between genetic markers and a given phenotype.
This thesis considers a situation where the phenotype is the absence or presence of
a common disease, the genetic marker is a biallelic single nucleotide polymorphism
(SNP), and environmental covariates are available. The main goal is to study and
compare four statistical methods (Score test, Likelihood ratio test, Wald test and
Cochran-Armitage test for trend) which, by using different approaches, test the
hypothesis about whether there is an association or not between the disease and
the genetic marker. The methods are applied to simulated datasets in order to
measure their test size and statistical power, and to compare them. Interaction
between the genetic marker and the environmental effect is also considered, and
strategies for simulating cohort and case-control data with genotype and environ-
mental covariates are studied.
The power simulations show that methods based on logistic regression models are
appropriate for detecting genotype-phenotype association, but when the environ-
mental effect is moderate, a simpler method (Cochran-Armitage test for trend)
which does not require model fitting at all, is adequate. When an interaction effect
is included in the model, the hypothesis testing becomes more complex. Several
possible approaches to this problem are discussed.
Place, publisher, year, edition, pages
Institutt for matematiske fag , 2013. , 84 p.
IdentifiersURN: urn:nbn:no:ntnu:diva-22628Local ID: ntnudaim:8972OAI: oai:DiVA.org:ntnu-22628DiVA: diva2:650404
Langaas, Mette, Førsteamanuensis