TABLE 5.

Dependence of population-attributable risk (PAR) on SNP frequencies in three studies of type II diabetes (Scott et al. 2007; Sladek et al. 2007; Zeggini et al. 2007)

SNP	Source	pcontrol	pcase	PAR
rs1801282	Scott et al. (2007)	0.823	0.848	0.16447
rs4402960	Scott et al. (2007)	0.304	0.341	0.056146
rs7754840	Scott et al. (2007)	0.36	0.387	0.044046
rs13266634	Scott et al. (2007)	0.609	0.649	0.11396
rs10811661	Scott et al. (2007)	0.85	0.872	0.17187
rs1111875	Scott et al. (2007)	0.522	0.546	0.052863
rs7903146	Scott et al. (2007)	0.181	0.227	0.059508
rs5219	Scott et al. (2007)	0.464	0.489	0.048924
rs9300039	Scott et al. (2007)	0.892	0.924	0.42105
rs8050136	Scott et al. (2007)	0.381	0.406	0.042088
rs7923837	Sladek et al. (2007)	0.623	0.665	0.11141
rs7480010	Sladek et al. (2007)	0.301	0.336	0.050072
rs3740878	Sladek et al. (2007)	0.728	0.76	0.11765
rs11037909	Sladek et al. (2007)	0.729	0.76	0.11439
rs1094369	Zeggini et al. (2007)	0.319	0.361	0.060957

Data from Scott et al. are from their Table S1; data from Sladek et al. are from their Table 1; data from Zeggini et al. are from their Table S2. When the same SNP was found in two or more studies, data from only one study were used. PAR is computed using Equation 4 in the text, with p_case = p^C and p_control = p.