ModularBoost: an efficient network inference algorithm based on module decomposition

BMC Bioinformatics

Table 2 AUROC and AUPR indexes for curated and PIDC networks with conventional GRN inference and ModularBoost

Methods	Curated GSD-1		Curated GSD-50		Curated GSD-70
Methods	AUROC	AUPR	AUROC	AUPR	AUROC	AUPR
Ridge	0.545	0.234	0.517	0.232	0.530	0.234
Linear regression	0.520	0.225	0.507	0.220	0.467	0.204
TIGRESS	0.547	0.233	0.548	0.253	0.523	0.249
GRNBoost2	0.547	0.230	0.566	0.258	0.550	0.258
ModularBoost	0.549	0.234	0.559	0.260	0.553	0.259

Methods	PIDC E. coli-S		PIDC E. coli-LL		PIDC E. coli-LH
Methods	AUROC	AUPR	AUROC	AUPR	AUROC	AUPR
Ridge	0.667	0.065	0.594	0.024	0.540	0.015
Linear regression	0.461	0.010	0.518	0.012	0.479	0.011
TIGRESS	0.676	0.068	0.612	0.035	0.601	0.020
GRNBoost2	0.659	0.056	0.558	0.038	0.570	0.034
ModularBoost	0.678	0.074	0.624	0.039	0.618	0.028

ISSN: 1471-2105