SNP Data

These tables give dbSNP mappings, ascertainment information, and NCBI Build 34 coordinates for the genotyped SNPs. Alleles are given for the (+) strand on the specified NCBI sequence. The tables give allele frequencies in the three genotyped population samples, and estimates of FST between each pair and jointly across all three population samples.

[02-Mar-06] The two-population Fst values have been updated so that values for positions that are not polymorphic are empty. Previously, these were reported as being zero. The updated values are consistent with Weir and Cockerham (1984), which says that Fst is undefined in these cases.

Genotype Data

These tables contain genotype results for 71 individuals: 23 African Americans, 24 European Americans, and 24 Han Chinese. For convenience, the Build 34 position information is also repeated here. Male genotypes at sex linked loci are represented here as homozygous diploid values but should be interpreted as haploid genotypes.

Linkage Disequilibrium Map Data

These tables describe results of the linkage disequilibrium bin analyses. Within each population sample, SNPs were grouped into bins of high LD, where at least one tagging SNP has r2 > 0.8 with every other SNP in the bin, using the algorithm of Carlson et al. (2004). Within each bin, the table identifies all SNPs that satisfy the tagging condition; to determine a minimal set of tagging SNPs, one such SNP would be selected for each bin. For each population sample, only the SNPs with MAF > 0.1 were grouped into bins.

Phased Haplotype Data

These tables contain inferred whole-chromosome haplotype results for each genotyped sample, generated with the HAP program (Halperin and Eskin, 2004). The form is similar to the genotype data tables however each column of diploid genotypes is split into two columns of haploid genotypes, where each column represents an inferred chromosome. Haplotype reconstruction was only performed for SNPs that had known chromosomal locations. The inferred haplotypes are expected to be locally very accurate, however, phasing errors are expected to be present at low frequency. As with the diploid data, males at sex linked loci are represented in these tables with homozygous diploid genotypes, and hence will have two identical haplotypes for regions that are actually haploid.

Haplotype Map Data

These tables describe results of partitioning the inferred haplotype data into .haplotype blocks. of limited diversity. Blocks were defined as sets of SNPs for which at least 80% of the inferred haplotypes could be grouped into common patterns with population frequencies of at least 5%. Haplotype block boundaries are very sensitive to the partitioning algorithm parameters and this represents just one of many possible partitioning schemes.

Nonsynonymous SNPs

This table describes 9370 SNPs that give rise to nonsynonymous substitutions in known genes annotated in NCBI Build 34 release 3. There are 10874 rows of data; the same SNP may appear several times if it appears in several overlapping transcripts.

dbSNP Mappings

This table describes the mapping between Perlegen snp local identifiers to NCBI dbSNP build 123 ss_id's and rs_id's.