Premium
A new multimarker test for family‐based association studies
Author(s) -
Rakovski Cyril S.,
Xu Xin,
Lazarus Ross,
Blacker Deborah,
Laird Nan M.
Publication year - 2007
Publication title -
genetic epidemiology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.301
H-Index - 98
eISSN - 1098-2272
pISSN - 0741-0395
DOI - 10.1002/gepi.20186
Subject(s) - single nucleotide polymorphism , snp , haplotype , association test , multiple comparisons problem , false discovery rate , genetic association , computer science , computational biology , genetics , biology , statistics , gene , mathematics , allele , genotype
We propose a new multimarker test for family‐based studies in candidate genes. We use simulations under different genetic models to assess the performance of competing testing strategies, characterized in this study as combinations of the following important factors: genes, statistical tests, tag single nucleotide polymorphisms (SNP) methods, number of tag SNPs and family designs. An ANOVA model is employed to provide descriptive summaries of the effects on power of the above‐mentioned factors. We find that tag SNP methods, gene characteristics and family designs have minimal impact on the best testing strategy. The familywise error rate (FWER) controlling multiple comparison procedure and the new multimarker test offer the highest power followed by the asymptotic global haplotype test.Both the FWER and the multimarker test are invariant to family designs and gain power as we increase the number of tag SNPs. However, the performance of the global haplotype test is slightly degraded when analyzing larger numbers of tag SNPs. Within the framework of our study, the best strategy for family‐based studies in candidate genes that emerged from our analysis is to use the FWER or the multimarker test and select 6–10 tag SNPs using any of the tag SNP methods considered. We confirm the conclusions of our study with an application to Alzheimer's disease data. Genet. Epidemiol. © 2006 Wiley‐Liss, Inc.