Phenotype harmonization and cross-study collaboration in GWAS consortia: the GENEVA experience

Abstract

Genome-wide association study (GWAS) consortia and collaborations formed to detect genetic loci for common phenotypes or investigate gene-environment (G*E) interactions are increasingly common. While these consortia effectively increase sample size, phenotype heterogeneity across studies represents a major obstacle that limits successful identification of these associations. Investigators are faced with the challenge of how to harmonize previously collected phenotype data obtained using different data collection instruments which cover topics in varying degrees of detail and over diverse time frames. This process has not been described in detail. We describe here some of the strategies and pitfalls associated with combining phenotype data from varying studies. Using the Gene Environment Association Studies (GENEVA) multi-site GWAS consortium as an example, this paper provides an illustration to guide GWAS consortia through the process of phenotype harmonization and describes key issues that arise when sharing data across disparate studies. GENEVA is unusual in the diversity of disease endpoints and so the issues it faces as its participating studies share data will be informative for many collaborations. Phenotype harmonization requires identifying common phenotypes, determining the feasibility of cross-study analysis for each, preparing common definitions, and applying appropriate algorithms. Other issues to be considered include genotyping timeframes, coordination of parallel efforts by other collaborative groups, analytic approaches, and imputation of genotype data. GENEVA's harmonization efforts and policy of promoting data sharing and collaboration, not only within GENEVA but also with outside collaborations, can provide important guidance to ongoing and new consortia.

Figure 1. Phenotype harmonization roles and responsibilities in GENEVA

Figure 2

Power of a case-control study to detect a gene-environment interaction (departure from a multiplicative odds model) when the binary exposure is measured perfectly or via a good proxy with 77% specificity and 99% sensitivity (roughly analogous to self-reported versus measured overweight status) This figure illustrates several points: a) large samples sizes are needed to detect gene-environment interactions; b) even modest misclassification can greatly decrease the power of tests for gene-environment interaction (and the relative decrease is greater for rare exposures); yet c) a large study using the proxy can have greater power than a smaller study using the perfect measure. This last point is important when the perfect measure is prohibitively expensive or only available on a small fraction of samples, while the good measure is relatively inexpensive or already available on many samples. Power calculations were performed using the methods described in Lindstrom et al. (2009), assuming a rare disease (prevalence 1 in 1,000), no main effect for the binary genetic factor (with 20% prevalence), an odds ratio of 1.5 for the exposure, an interaction odds ratio of 1.35, and a Type I error rate of 5×10^-8.