Methods for Investigating Gene-Environment Interactions in Candidate Pathway and Genome-Wide Association Studies

Department of Preventive Medicine, University of Southern California, Los Angeles, California, 90089-9011, USA.
Annual Review of Public Health (Impact Factor: 6.47). 04/2010; 31(1):21-36. DOI: 10.1146/annurev.publhealth.012809.103619
Source: PubMed


Despite the considerable enthusiasm about the yield of novel and replicated discoveries of genetic associations from the new generation of genome-wide association studies (GWAS), the proportion of the heritability of most complex diseases that have been studied to date remains small. Some of this "dark matter" could be due to gene-environment (G x E) interactions or more complex pathways involving multiple genes and exposures. We review the basic epidemiologic study design and statistical analysis approaches to studying G x E interactions individually and then consider more comprehensive approaches to studying entire pathways or GWAS data. In addition to the usual issues in genetic association studies, particular care is needed in exposure assessment, and very large sample sizes are required. Although hypothesis-driven, pathway-based and agnostic GWA study approaches are generally viewed as opposite poles, we suggest that the two can be usefully married using hierarchical modeling strategies that exploit external pathway knowledge in mining genome-wide data.

Download full-text


Available from: Duncan C Thomas
  • Source
    • "There are a growing number of literature works on methods for pathway modeling, motivated in large part by an interest in mining GWAS data for commonalities across related genes that individually may not achieve genomewide significance but in the aggregate may point to novel pathways (see [18] for a review of gene set enrichment analysis and alternatives). Our goal here is more modest, guided by an a priori selection of strong candidate genes [19]. Like other methods of pathway analysis, however, we aim to exploit external knowledge about the biological function of each gene and the relationships between them [20]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: A variety of methods have been proposed for studying the association of multiple genes thought to be involved in a common pathway for a particular disease. Here, we present an extension of a Bayesian hierarchical modeling strategy that allows for multiple SNPs within each gene, with external prior information at either the SNP or gene level. The model involves variable selection at the SNP level through latent indicator variables and Bayesian shrinkage at the gene level towards a prior mean vector and covariance matrix that depend on external information. The entire model is fitted using Markov chain Monte Carlo methods. Simulation studies show that the approach is capable of recovering many of the truly causal SNPs and genes, depending upon their frequency and size of their effects. The method is applied to data on 504 SNPs in 38 candidate genes involved in DNA damage response in the WECARE study of second breast cancers in relation to radiotherapy exposure.
    Full-text · Article · Dec 2013 · International Journal of Genomics
  • Source
    • "In a simple case-control study when the outcome is a binary variable, exploration of genotype, and gender interaction by logistic regression is a mostly neglected, but rather straightforward approach. For more complicated designs and for quantitative outcomes, methods developed to analyze gene and environment interactions can be used as if gender was a binary environmental exposure variable (Thomas, 2010). "
    [Show abstract] [Hide abstract]
    ABSTRACT: The gender difference in cancer susceptibility is one of the most consistent findings in cancer epidemiology. Hematologic malignancies are generally more common in males and this can be generalized to most other cancers. Similar gender differences in non-malignant diseases including autoimmunity, are attributed to hormonal or behavioral differences. Even in early childhood, however, where these differences would not apply, there are differences in cancer incidence between males and females. In childhood, few cancers are more common in females, but overall, males have higher susceptibility. In Hodgkin lymphoma, the gender ratio reverses toward adolescence. The pattern that autoimmune disorders are more common in females, but cancer and infections in males suggests that the known differences in immunity may be responsible for this dichotomy. Besides immune surveillance, genome surveillance mechanisms also differ in efficiency between males and females. Other obvious differences include hormonal ones and the number of X chromosomes. Some of the differences may even originate from exposures during prenatal development. This review will summarize well-documented examples of gender effect in cancer susceptibility, discuss methodological issues in exploration of gender differences, and present documented or speculated mechanisms. The gender differential in susceptibility can give important clues for the etiology of cancers and should be examined in all genetic and non-genetic association studies.
    Full-text · Article · Nov 2012 · Frontiers in Genetics
  • Source
    • "Observed human variation in response to environmental stressors suggests the possibility of gene by environment (G × E) interactions (Gottlieb, 2003; Johnston and Edwards, 2002). Identifying these G × E interactions through empirical research, however, has been difficult (Burmeister et al., 2008; Thomas, 2010b). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Here we examine the effects of both self-reported and independent observer-reported environmental risk indices, the serotonin transporter gene promoter (5HTTLPR) polymorphism, and their interaction on self-esteem. This trait was assessed during early and mid adolescence (mean age=14 and 16.5, respectively) and young adulthood (mean age=21.8) in a prospective cohort of 1214 unrelated participants in the Longitudinal Study of Adolescent Health (Add Health). Using structural equation modeling we identified a gene-environment (G×E) interaction using observer-report but not self-report measures of environmental stress exposure during adolescence: 5HTTLPR genotype and observer-reports of home and neighborhood quality (HNQ) during adolescence interacted to predict self-esteem levels in young adulthood (p<.004). Carriers of the s allele who lived in poor HNQ conditions during adolescence reported lower self-esteem in young adulthood than those with a good HNQ during adolescence. In contrast, among individuals with the l/l genotype, adolescent HNQ did not predict adulthood self-esteem. Genes may moderate the effect of adolescent environmental conditions on adulthood self-esteem.
    Full-text · Article · May 2012 · Biological psychology
Show more