Genome-wide association studies

What are we to make of genome-wide association studies (GWAS)? In an editorial highlighting 3 articles in this month’s issue (Wu et al.; Li et al.; and Mathias et al.), Donata Vercelli, MD, explores this question. Dr. Vercelli points out the advantage that GWAS offer over linkage and candidate gene studies; namely, that GWAS can look for susceptibility SNPs across a whole genome without a priori assumptions. That’s also the difficulty, because GWAS are now uncovering SNP associations from surprising gene families, and biological validation will be required for these results. Additionally, currently available GWAS technologies may contribute to variability of results because they do not account for rare SNPs and admixed populations. Finally, Dr. Vercelli talks about the bugbear of epigenetic influence, which is difficult to account for in GWAS.

As for the articles themselves:

Mathias et al. look at genome-wide associations in asthmatic and non-asthmatic African American and African Caribbean populations. They identify only 3 SNPs in 3 genes with significant association from combined analysis of the results of GWAS of each population. The role of the genes they single out–ADRA1B, PRNP, and DPP10–in asthma susceptibility is not yet understood. The authors also report that their results did not generalize to 2 white populations. They also stress the difficulties inherent to studying non-Caucasian populations by using genotyping platforms designed to assess variants common in Caucasians.

Li et al. investigate SNP associations in severe and difficult-to-treat, white, asthmatic populations using 14, highly replicated, candidate genes. Li and co-authors show that SNPs in two regions, RAD50-IL13 and HLA-DR/DQ, have consistent association with asthma, with SNPs in the RAD50 region having the strongest association. While the former regions have functional impact on Th2 cytokine expression and antigen-presentation respectively, RAD50 is involved in DNA repair and its effect on asthma susceptibility isn’t known. Li et al. propose that a RAD50 region characterized in mice as affecting Th cytokine expression may have a functional role in humans. The authors suggest that variants in the RAD50locus should be considered new asthma candidates.

Wu et al. report results of a GWA study that focused on previously published asthma candidate genes in the hope that this approach would improve signal detection. Out of 118 asthma candidate genes, 4 with multiple SNPs had significant associations in the pediatric asthma/Mexican population studied: TGFB1, IL1RL1, IL18R1, and DPP10.

Do you have any questions or comments about these studies? We want to hear from you. Please post your questions and comments below.