Methods for meta-analyses of genome-wide association studies: critical assessment of empirical evidence

Am J Epidemiol. 2012 Apr 15;175(8):739-49. doi: 10.1093/aje/kwr385. Epub 2012 Mar 16.


There has been a steep increase in the number of meta-analyses of genome-wide association (GWA) studies aimed at identifying genetic variants with increasingly smaller effects, but pressure to publish findings of new genetic associations has limited the time available for careful consideration of all of their methodological aspects. The authors surveyed the literature (2007-2010) to provide empirical evidence on the methods used in GWA meta-analyses, including their organization, requirements about the uniformity of methods used in primary studies, methods for data pooling, investigation of between-study heterogeneity, and quality of reporting. This review showed that a great variety of methods are being used, but the rationale for their choice is often unclear. It also highlights how important methodological aspects have received insufficient attention, potentially leading to missed opportunities for improving gene discovery and characterization. Evaluation of power to replicate findings was inadequate, and the number of variants selected for replication was not associated with replication sample size. A low proportion of GWA meta-analyses investigated the presence and magnitude of heterogeneity, even when there was little uniformity in methods used in primary studies. More methodological work is required before clear guidance can be offered as to optimal methods or tradeoffs between alternative methods. However, there is a clear need for guidelines for reporting the results of GWA meta-analyses.

Publication types

  • Review

MeSH terms

  • Data Interpretation, Statistical
  • Genome-Wide Association Study* / methods
  • Genome-Wide Association Study* / statistics & numerical data
  • Guidelines as Topic
  • Humans
  • Meta-Analysis as Topic*
  • Polymorphism, Single Nucleotide
  • Research Report
  • Sample Size