Evaluating the power to discriminate between highly correlated SNPs in genetic association studies

Genet Epidemiol. 2010 Jul;34(5):463-8. doi: 10.1002/gepi.20504.

Abstract

Neighboring common polymorphisms are often correlated (in linkage disequilibrium (LD)) as a result of shared ancestry. An association between a polymorphism and a disease trait may therefore be the indirect result of a correlated functional variant, and identifying the true causal variant(s) from an initial disease association is a major challenge in genetic association studies. Here, we present a method to estimate the sample size needed to discriminate between a functional variant of a given allele frequency and effect size, and other correlated variants. The sample size required to conduct such fine-scale mapping is typically 1-4 times larger than required to detect the initial association. Association studies in populations with different LD patterns can substantially improve the power to isolate the causal variant. An online tool to perform these calculations is available at http://moya.srl.cam.ac.uk/ocac/FineMappingPowerCalculator.html.

MeSH terms

  • Alleles
  • Breast Neoplasms / genetics
  • Female
  • Genetic Predisposition to Disease*
  • Genetic Variation
  • Genotype
  • Haplotypes
  • Humans
  • Internet
  • Linkage Disequilibrium
  • Models, Genetic*
  • Models, Statistical*
  • Polymorphism, Single Nucleotide / genetics*
  • Receptor, Fibroblast Growth Factor, Type 2 / genetics
  • Sample Size

Substances

  • FGFR2 protein, human
  • Receptor, Fibroblast Growth Factor, Type 2