Finding and using haplotype blocks in candidate gene association studies

Abstract Recent discoveries (Daly et al. , ; Jeffreys et al. , ; Gabriel et al. , ) that there is considerable heterogeneity in the amount of linkage disequilibrium (LD) between single‐nucleotide polymorphism (SNP) markers over the genome and the implication that recombination rates themselves may also be considerably heterogeneous suggest specific analysis strategies for design and analysis of genetic association studies of disease. This article focuses upon the implications for such studies of the existence of well‐defined “haplotype blocks” covering much of the genome over which there is little evidence for recombination and within which there is limited haplotype diversity. The main topics addressed are: SNP and haplotype discovery Picking of haplotype‐tagging SNPs (ht SNPs) for large‐scale genotyping (e.g., in a case‐control study) Testing and estimation of haplotype‐specific risk in a large‐scale case–control study.