We selected ten 100-kb regions for direct PCR-Sanger capillary sequencing analysis. These regions included the central 100 kb from five previously sequenced HapMap-ENCODE 500-kb regions12 and five ENCODE regions not previously subject to sequencing in the HapMap Project (Supplementary Table 4). A total of 692 unrelated samples chosen from the ten then available genotyped population samples (ASW, CEU, CHB, CHD, GIH, JPT, LWK, MXL, TSI and YRI) were interrogated and passed quality control metrics (Supplementary Table 1). SNPs were discovered from the raw sequence data using SNP Detector 3.0 software13. Subsequent genotyping showed an overall genotype concordance rate of 99.2% and an 86.8% genotype concordance rate for genotypes with minor alleles (Supplementary Table 5a). Also, a 93.6% genotype concordance rate was found for singleton genotypes with minor alleles and 88% for two to six copies of the minor allele. The higher genotype concordance rate in singletons reflects the higher stringency applied in making singleton calls. (See Supplementary Information and Supplementary Table 5 for details.)
