Perspectives on human genetic variation from the HapMap Project

Abstract

The completion of the International HapMap Project marks the start of a new phase in human genetics. The aim of the project was to provide a resource that facilitates the design of efficient genome-wide association studies, through characterising patterns of genetic variation and linkage disequilibrium in a sample of 270 individuals across four geographical populations. In total, over one million SNPs have been typed across these genomes, providing an unprecedented view of human genetic diversity. In this review we focus on what the HapMap Project has taught us about the structure of human genetic variation and the fundamental molecular and evolutionary processes that shape it.

Figure 1. The Relationship between Genealogical History and Allelic Association

The upper part of the figure represents the genealogy for the 13 haplotypes observed in a 40-kb region of Chromosome 1 (between SNPs rs12085605 and rs932087) where there is no evidence for recombination (for no pair of SNPs are all four possible combinations of alleles observed), with the location of polymorphic mutations indicated by circles. The lower part of the figure indicates the relative frequency of each haplotype in the sample from each of the three panels (in greyscale, with white indicating 0% and black indicating 100%). The dotted line in the genealogy indicates a branch of the tree that is not present in the CEU sample and whose removal results in perfect association between SNPs rs12085824 and rs11205476.

Figure 2. Patterns of Haplotype Structure and Recombination in the HapMap ENCODE Region on Chromosome 7q31.33

The estimated recombination rate (in centimorgans per megabase) is shown as a dark blue line, with statistically significant recombination hotspots (see [15] for details) as grey lines. For each analysis panel, each non-redundant haplotype with a frequency of at least 10% is represented by a horizontal line between the starting and ending SNPs (see [15] for details of methodology); the vertical height of these lines is arbitrary. Note that only one of the six hotspots is sufficiently strong to break all common haplotypes.

Figure 3. The Relationship between Recombination Rate, Recombination Hotspots, and the Location of Untaggable SNPs

For two HapMap ENCODE regions the estimated recombination rate (dark blue line) and the location of statistically significant hotspots (grey lines) are shown along with the location of SNPs that are untaggable in the YRI (green) CEU (red), or CHB + JPT (purple) panels. Note that most, but not all, untaggable SNPs occur in recombination hotspots.