Multipoint linkage-disequilibrium mapping narrows location interval and identifies mutation heterogeneity.

Journal article

Single-nucleotide polymorphism (SNP) genotypes were recently examined in an 890-kb region flanking the human gene CYP2D6. Single-marker and haplotype-based analyses identified, with genomewide significance (P < 10-7), a 403-kb interval displaying strong linkage disequilibrium (LD) with predicted poor-metabolizer phenotype. However, the width of this interval makes the location of causal variants difficult: for example, the interval contains seven known or predicted genes in addition to CYP2D6. We have developed the Bayesian fine-mapping software coldmap, which, applied to these genotype data, yields a 95% location interval covering only 185 kb and establishes genomewide significance for a causal locus within the region. Strikingly, our interval correctly excludes four SNPs, which individually display association with genomewide significance, including the SNP showing strongest LD (P < 10-34). In addition, coldmap distinguishes homozygous cases for the major CYP2D6 mutation from those bearing minor mutations. We further investigate a selection of SNP subsets and find that previously reported methods lead to a 38% savings in SNPs at the cost of an increase of <20% in the width of the location interval.

Authors: Morris, A; Whittaker, J; Xu, C; Hosking, L; Balding, D

• Publication status: Published
• Journal: Proceedings of the National Academy of Sciences of the United States of America
• Volume: 100
• Issue: 23
• Pages: 13442-13446
• Publication date: 2003-11-01
• DOI: 10.1073/pnas.2235031100
• EISSN: 1091-6490
• ISSN: 0027-8424
• Language: English
• Keywords: Phenotype; Models, Genetic; Bayes Theorem; Polymorphism, Single Nucleotide; Mutation; Cytochrome P-450 CYP2D6; Humans; Genotype; Phylogeny; Software; Linkage Disequilibrium; Chromosome Mapping