Table 3.

Description of the HapMap and the 1000 Genomes Project gold standards used in this study

	HapMap			1000GP Whole Genome
Sample	Positions	Variants	Heterozygous variants	Positions	Variants	Heterozygous variants
CEU-D	42,964	11,558	6,568	24,926,557	14,605	9,595
CEU-M	42,967	11,455	6,460	7,038,292	3,489	2,514
CEU-F	43,049	11,461	6,498	12,227,792	6,030	4,012
YRI-D	43,161	12,320	7,041	25,818,250	18,730	12,569
YRI-M	43,219	12,205	6,843	2,356,922	1,136	831
YRI-F	43,246	12,202	6,734	11,322,826	6,393	4,052

Both gold standards comprise a set of positions within CCDS for each of the six samples for which genotypes are given. We use the gold standards to compare the given genotypes to the genotype calls we make from our capture data over the same positions. The positions for the HapMap gold standard are taken to be the CCDS positions that have been successfully genotyped by the HapMap project using SNP arrays. The positions for the 1000 Genomes Project gold standard are the ones for which we were able to obtain high confidence (minimum consensus quality of 100) genotype calls based on the 1000 Genomes Project trio pilot sequence data. For both gold standards, the Positions column specifies the number of positions in the gold standard for each of the six samples, the Variants column specifies the number of gold standard genotypes that differ from the reference allele at the corresponding position, and the Heterozygous variants column specifies the number of heterozygous gold standard genotypes.