Identifying genetic variation affecting a complex trait in simulated data: a comparison of meta-analysis with pooled data analysis

BMC Genet. 2005 Dec 30;6 Suppl 1(Suppl 1):S97. doi: 10.1186/1471-2156-6-S1-S97.

Abstract

We explored the power and consistency to detect linkage and association with meta-analysis and pooled data analysis using Genetic Analysis Workshop 14 simulated data. The first 10 replicates from Aipotu population were used. Significant linkage and association was found at all 4 regions containing the major loci for Kofendrerd Personality Disorder (KPD) using both combined analyses although no significant linkage and association was found at all these regions in a single replicate. The linkage results from both analyses are consistent in terms of the significance level of linkage test and the estimate of locus location. After correction for multiple-testing, significant associations were detected for the same 8 single-nucleotide polymorphisms (SNP) in both analyses. There were another 2 SNPs for which significant associations with KPD were found only by pooled data analysis. Our study showed that, under homogeneous condition, the results from meta-analysis and pooled data analysis are similar in both linkage and association studies and the loss of power is limited using meta-analysis. Thus, meta-analysis can provide an overall evaluation of linkage and association when the original raw data is not available for combining.

Publication types

  • Comparative Study
  • Research Support, N.I.H., Extramural

MeSH terms

  • Computer Simulation*
  • Databases, Genetic*
  • Family
  • Genetic Predisposition to Disease
  • Genetic Variation*
  • Humans
  • Meta-Analysis as Topic*
  • Physical Chromosome Mapping
  • Quantitative Trait, Heritable*