Table 1.

Estimated proportions of variance in disease liability explained by SNPs.

Disease	Cases	Controls	$h_g^2$ (s.e.)
Schizophrenia	22,177	27,629	0.274 (0.007)
Allergic rhinitis	13,437	41,297	0.074 (0.015)
Asthma	8,929	45,805	0.152 (0.018)
Cardiovasc. dis.	14,861	39,873	0.092 (0.015)
Diabetes type 2	6,845	47,889	0.297 (0.022)
Dyslipidemia	29,511	25,223	0.263 (0.014)
Hypertension	27,921	26,813	0.255 (0.014)
Macular degen.	3,700	51,034	0.242 (0.029)
Osteoarthritis	19,832	34,902	0.098 (0.014)
Osteoporosis	5,337	49,397	0.195 (0.024)

Schizophrenia cases and controls are from the PGC2 data set¹²; the $h_g^2$ estimate assumes a population risk of 1% and was computed using PCGC regression to avoid REML bias induced by over-ascertainment of cases^17,27. Cases and controls for the other 9 diseases are from the GERA data set; $h_g^2$ estimates assume random sample ascertainment and were computed using BOLT-REML.