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

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.     

Sifting the wheat from the chaff: prioritizing GWAS results by identifying consistency across analytical methods

The curse of multiple testing has led to the adoption of a stringent Bonferroni threshold for declaring genome-wide statistical significance for any one SNP as standard practice. Although justified in avoiding false positives, this conservative approach has the potential to miss true associations as most studies are drastically underpowered. As an alternative to increasing sample size, we compare results from a typical SNP-by-SNP analysis with three other methods that incorporate regional information in order to boost or dampen an otherwise noisy signal: the haplotype method (Schaid et al. [2002] Am J Hum Genet 70:425-434), the gene-based method (Liu et al. [2010] Am J Hum Genet 87:139-145), and a new method (interaction count) that uses genome-wide screening of pairwise SNP interactions. Using a modestly sized case-control study, we conduct a genome-wide association studies (GWAS) of age-related macular degeneration, and find striking agreement across all methods in regions of known associated variants. We also find strong evidence of novel associated variants in two regions (Chromosome 2p25 and Chromosome 10p15) in which the individual SNP P-values are only suggestive, but where there are very high levels of agreement between all methods. We propose that consistency between different analysis methods may be an alternative to increasingly larger sample sizes in sifting true signals from noise in GWAS.     

A Nonparametric Test to Detect Quantitative Trait Loci Where the Phenotypic Distribution Differs by Genotypes

Searching for genetic variants involved in gene‐gene and gene‐environment interactions in large‐scale data raises multiple methodological issues. Many existing methods have focused on the problem of dimensionality, trying to explore the largest number of combinations between risk factors while considering simple interaction models. Despite evidence demonstrating the efficacy of these methods in simulated data, their application in real data has been unsuccessful so far. The classical test of a linear marginal genetic effect has been widely used for agnostic genome‐wide association studies, with the underlying idea that most variants involved in interactions might display marginal effect on the phenotypic mean. Although this approach may allow for the identification of genetic variants involved in interactions in many scenarios, the linear marginal effects of some causal alleles on the phenotypic mean might not be always detectable at genome‐wide significance level. We introduce in this study a general association test for quantitative trait loci that compare the distributions of phenotypic values by genotypic classes as opposed to most standard tests that compare phenotypic means by genotypic classes. Using simulations we show that in presence of interactions, this approach can be more powerful than the standard test of the linear marginal effect, with a gain of power increasing with increasing interaction effect and decreasing frequencies of the interacting exposures. We demonstrate the potential utility of our method on real data by analyzing mammographic density genome‐wide data from the Nurses’ Health Study.     

Examining the statistical properties of fine-scale mapping in large-scale association studies

Interpretation of dense single nucleotide polymorphism (SNP) follow-up of genome-wide association or linkage scan signals can be facilitated by establishing expectation for the behaviour of primary mapping signals upon fine-mapping, under both null and alternative hypotheses. We examined the inferences that can be made regarding the posterior probability of a real genetic effect and considered different disease-mapping strategies and prior probabilities of association. We investigated the impact of the extent of linkage disequilibrium between the disease SNP and the primary analysis signal and the extent to which the disease gene can be physically localised under these scenarios. We found that large increases in significance (>2 orders of magnitude) appear in the exclusive domain of genuine genetic effects, especially in the follow-up of genome-wide association scans or consensus regions from multiple linkage scans. Fine-mapping significant association signals that reside directly under linkage peaks yield little improvement in an already high posterior probability of a real effect. Following fine-mapping, those signals that increase in significance also demonstrate improved localisation. We found local linkage disequiliptium patterns around the primary analysis signal(s) and tagging efficacy of typed markers to play an important role in determining a suitable interval for fine-mapping. Our findings help inform the interpretation and design of dense SNP-mapping follow-up studies, thus facilitating discrimination between a genuine genetic effect and chance fluctuation (false positive).     

On the follow-up of genome-wide association studies: an overall test for the most promising SNPs

Even in large-scale genome-wide association studies (GWASs), only a fraction of the true associations are detected at the genome-wide significance level. When few or no associations reach the significance threshold, one strategy is to follow up on the most promising candidates, i.e. the single nucleotide polymorphisms (SNPs) with the smallest association-test P-values, by genotyping them in additional studies. In this communication, we propose an overall test for GWASs that analyzes the SNPs with the most promising P-values simultaneously and therefore allows an early assessment of whether the follow-up of the selected SNPs is likely promising. We theoretically derive the properties of the proposed overall test under the null hypothesis and assess its power based on simulation studies. An application to a GWAS for chronic obstructive pulmonary disease suggests that there are true association signals among the top SNPs and that an additional follow-up study is promising.     

Tests for gene-environment interaction from case-control data: a novel study of type I error, power and designs

To evaluate the risk of a disease associated with the joint effects of genetic susceptibility and environmental exposures, epidemiologic researchers often test for non-multiplicative gene-environment effects from case-control studies. In this article, we present a comparative study of four alternative tests for interactions: (i) the standard case-control method; (ii) the case-only method, which requires an assumption of gene-environment independence for the underlying population; (iii) a two-step method that decides between the case-only and case-control estimators depending on a statistical test for the gene-environment independence assumption and (iv) a novel empirical-Bayes (EB) method that combines the case-control and case-only estimators depending on the sample size and strength of the gene-environment association in the data. We evaluate the methods in terms of integrated Type I error and power, averaged with respect to varying scenarios for gene-environment association that are likely to appear in practice. These unique studies suggest that the novel EB procedure overall is a promising approach for detection of gene-environment interactions from case-control studies. In particular, the EB procedure, unlike the case-only or two-step methods, can closely maintain a desired Type I error under realistic scenarios of gene-environment dependence and yet can be substantially more powerful than the traditional case-control analysis when the gene-environment independence assumption is satisfied, exactly or approximately. Our studies also reveal potential utility of some non-traditional case-control designs that samples controls at a smaller rate than the cases. Apart from the simulation studies, we also illustrate the different methods by analyzing interactions of two commonly studied genes, N-acetyl transferase type 2 and glutathione s-transferase M1, with smoking and dietary exposures, in a large case-control study of colorectal cancer.