Table 1. Mean CORs and Regression slopes of infinitesimal and non-infinitesimal methods in independent validation cohort of CE, CEL, and T2D.

For two-trait prediction models, we jointly modeled CD with UC, CEL with UC and, T2D with CAD.

	CORa			Regression Slopeb
	CD	CEL	T2D	CD	CEL	T2D
ldpred-inf	0.196	0.072	0.137	0.454	0.168	1.99
AnnoPred-inf	0.219	0.098	0.145	0.572	0.255	2.15
PleioPred-inf	0.246	0.100	0.168	0.661	0.292	2.198
PleioPred-anno-inf	0.248	0.122	0.184	0.739	0.400	2.333
ldpred	0.247	0.120	0.217	0.873	0.661	2.83
AnnoPred	0.279	0.132	0.219	1.306	0.924	2.86
PleioPred	0.307	0.141	0.225	1.284	1.332	3.05
PleioPred-anno	0.297	0.156	0.22	1.340	1.361	3.063

^a correlations between disease status and PRS;

^b Regression slopes of logistic regression with case/control status as outcome and PRS as covariates, larger value indicates a larger increase in odds ratio when PRS increases by one unit.