Table 2.

Correlations of SMILE results with heritability estimates from published studies

Cohort	Scenario	N	H2 correlation between SMILE and published study (P-value)		Percentage of traits with overlapping conference Interval		Mean squared differences between SMILE and published estimates of H2		Median difference in H2 (SMILE – published estimates)
			GPC	SMILE	GPC	SMILE	GPC	SMILE	GPC	SMILE
CaTCH3	All	540	0.11 (0.014)	0.12 (0.0040)	0.43	0.42	0.060	0.053	0.002	−0.031
	K > 1%	405	0.19 (8.7 × 10−5)	0.21(2.3 × 10−5)	0.42	0.41	0.049	0.044	−0.013	−0.045
LDSC-UKB34,79**	All	68	0.18 (0.14)	0.16 (0.20)	0.75	0.79	0.11	0.099	0.189	0.169
	K > 1%	44	0.33 (0.029)	0.30 (0.050)	0.66	0.73	0.12	0.098	0.205	0.169
MS14	All	63	0.79 (1.7 × 10−14)	0.83 (2.7 × 10−17)	0.44	0.26	0.016	0.020	−0.062	−0.101
	K > 1%	52	0.76 (5.77 × 10−11)	0.79 (4.7 × 10−12)	0.41	0.2	0.017	0.021	−0.064	−0.106
NY15	All	33	0.57 (0.00051)	0.52 (0.0015)	0.7	0.65	0.029	0.032	−0.009	−0.049
	K > 1%	31	0.48 (0.0069)	0.48 (0.0069)	0.68	0.68	0.028	0.030	−0.009	−0.053

($K>0.01$ indicates traits with an observed prevalence greater than 1%.) All p-values are for two-sided hypothesis tests.

^**The NY and LDSC-UKB studies are based on ICD-9/10 codes. It is difficult to directly compare ICD code defined and PheWAS code defined phenotypes. For this reason, we restricted our comparisons to ICD-9/10 codes which could be mapped to a unique PheWAS code.