Table 1. Precision and gene overlap for various competing models for the task of phenotype discovery, evaluated using disease-gene associations.

Data	K	Average precision@K				Average perc. of overlapping genes@K
		1	2	5	10	1	2	5	10
	D2D	0.9449	0.9367	0.9225	0.9047	0.8159	0.7966	0.7564	0.7111
	Modularity (Adjacency)	0.8575	0.8457	0.8284	0.8130	0.5198	0.4893	0.4508	0.4145
	Spectral (Adjacency)	0.7181	0.7007	0.6795	0.6640	0.3052	0.2779	0.2311	0.2006
	Modularity (Comorbidity)	0.8493	0.8412	0.8110	0.7865	0.5586	0.5315	0.4681	0.4204
	Spectral (Comorbidity)	0.8375	0.8316	0.8190	0.7974	0.5288	0.4989	0.4520	0.3964
	Comorbidity graph	0.7268	0.7134	0.6915	0.7068	0.1582	0.1496	0.1465	0.1554
	Disease co-occurrence	0.5616	0.5516	0.5439	0.5668	0.1448	0.1329	0.1264	0.1261
	LDA	0.5260	0.5094	0.4913	0.4217	0.1040	0.0989	0.0864	0.0853
	DAG2D	0.9598	0.9444	0.9239	0.9079	0.8486	0.7963	0.7237	0.6720
	Modularity (Adjacency)	0.8711	0.8604	0.8503	0.8389	0.5303	0.5082	0.4706	0.4340
	Spectral (Adjacency)	0.9165	0.9102	0.9020	0.8926	0.7524	0.7430	0.7277	0.7110
	Disease and genes co-occurrence	0.6978	0.6985	0.7093	0.7071	0.1058	0.1042	0.1018	0.0935
	LDA	0.5795	0.3874	0.3253	0.2831	0.1136	0.0781	0.0760	0.0652