Table 3.

GWAS Concordance for Type 2 Diabetes Exome Sequencing Data

$\hat{a}$	No. Samples	Ignored	Adjusted	Paired
Total Concordance
0%–1%	202	0.998	0.998	0.998
1%–5%	293	0.996	0.998	0.998
5%–10%	218	0.958	0.997	0.998
10%–15%	591	0.920	0.993	0.996
15%–20%	169	0.878	0.984	0.992
>20%	30	0.841	0.950	0.971
Total	1,503	0.945	0.993	0.996
Non-reference Concordance
0%–1%	202	0.996	0.997	0.997
1%–5%	293	0.992	0.995	0.995
5%–10%	218	0.908	0.993	0.994
10%–15%	591	0.833	0.985	0.991
15%–20%	169	0.760	0.964	0.983
>20%	30	0.702	0.890	0.936
Total	1,503	0.882	0.985	0.991
r2
0%–1%	202	0.997	0.998	0.998
1%–5%	293	0.994	0.996	0.996
5%–10%	218	0.929	0.995	0.996
10%–15%	591	0.863	0.990	0.994
15%–20%	169	0.791	0.977	0.989
>20%	30	0.725	0.930	0.946
Total	1,503	0.905	0.990	0.994

Mean per-sample genotype accuracy with the GWAS data when we ignore contamination, adjust without regard for the source of contamination, and adjust using known contamination source.