Figure - PMC

Skip to main content

View full-text article in PMC

. Author manuscript; available in PMC: 2012 Sep 30.

Published in final edited form as: Cell. 2011 Sep 30;147(1):57–69. doi: 10.1016/j.cell.2011.09.011

A. Common variants. This image illustrates the structure of common variants and linkage disequilibrium blocks. The top lists a reference genome spanning ~10 kb and the reference genotypes of the variants. The haplotype structure is broken up into two blocks by a recombination hotspot. Each block contains a set of markers in tight LD, which can be phased into a small number of haplotypes. Below that, a limited number of genotypes are depicted for a hypothetical individual because a commercial array would assay only a limited collection of all of the common variants in a region. The bottom row demonstrates how data for those genotypes can be phased using reference population data and how missing genotypes can be imputed if the haplotype can be inferred accurately. In some instances, imputed genotypes may not be uncertain. B. Fine mapping with conditional haplotype analysis. This figure illustrates the basic concept of conditional haplotype mapping. The left-hand side lists genotypes at ten variant sites (numbered) that define 7 common haplotypes. Each row represents a haplotype, and genotypes at variant sites are listed in each column. Assuming that a common variant association is observed at marker 1, identical associations will be observed at the markers 2, 3, and 5 because their genotypes are correlated across haplotypes. In the first step, haplotypes are grouped by marker 1. The result is that the seven haplotypes form two subgroups (indicated by purple and red bars on the right). One group demonstrates association with disease (right). Including marker 7 breaks the groups up further into four haplotypes (indicated by purple, green, blue, and red bars on the far right). By adding marker 7, differential risk association between haplotypes is apparent. Whereas the T/G haplotype confers risk, the T/T haplotype confers even more risk.

A. Common variants. This image illustrates the structure of common variants and linkage disequilibrium blocks. The top lists a reference genome spanning ~10 kb and the reference genotypes of the variants. The haplotype structure is broken up into two blocks by a recombination hotspot. Each block contains a set of markers in tight LD, which can be phased into a small number of haplotypes. Below that, a limited number of genotypes are depicted for a hypothetical individual because a commercial array would assay only a limited collection of all of the common variants in a region. The bottom row demonstrates how data for those genotypes can be phased using reference population data and how missing genotypes can be imputed if the haplotype can be inferred accurately. In some instances, imputed genotypes may not be uncertain. B. Fine mapping with conditional haplotype analysis. This figure illustrates the basic concept of conditional haplotype mapping. The left-hand side lists genotypes at ten variant sites (numbered) that define 7 common haplotypes. Each row represents a haplotype, and genotypes at variant sites are listed in each column. Assuming that a common variant association is observed at marker 1, identical associations will be observed at the markers 2, 3, and 5 because their genotypes are correlated across haplotypes. In the first step, haplotypes are grouped by marker 1. The result is that the seven haplotypes form two subgroups (indicated by purple and red bars on the right). One group demonstrates association with disease (right). Including marker 7 breaks the groups up further into four haplotypes (indicated by purple, green, blue, and red bars on the far right). By adding marker 7, differential risk association between haplotypes is apparent. Whereas the T/G haplotype confers risk, the T/T haplotype confers even more risk.