Graph embedding ensemble methods based on the heterogeneous network for lncRNA-miRNA interaction prediction

BMC Genomics

Table 3 Performances of based predictors and the ensemble models

Embedding	AUPR	AUC	F1	ACC	REC	SPEC	PRE
LE	0.6654 ± 0.0033	0.9430 ± 0.0017	0.6592 ± 0.0040	0.9976 ± 0.0001	0.5429 ± 0.0079	0.9995 ± 0.0001	0.8420 ± 0.0144
GraRep	0.6805 ± 0.0037	0.9417 ± 0.0019	0.6818 ± 0.0036	0.9977 ± 0.0001	0.5703 ± 0.0066	0.9996 ± 0.0001	0.8498 ± 0.0137
HOPE	0.6573 ± 0.0036	0.9281 ± 0.0022	0.6796 ± 0.0035	0.9976 ± 0.0001	0.5813 ± 0.0087	0.9994 ± 0.0001	0.8198 ± 0.0134
DeepWalk	0.6511 ± 0.0037	0.9383 ± 0.0018	0.6463 ± 0.0051	0.9974 ± 0.0001	0.5452 ± 0.0133	0.9994 ± 0.0001	0.7993 ± 0.0248
GAE	0.6664 ± 0.0031	0.9292 ± 0.0023	0.6754 ± 0.0033	0.9976 ± 0.0001	0.5666 ± 0.0086	0.9995 ± 0.0001	0.8395 ± 0.0185
GEEL-PI	0.7004 ± 0.0035	0.9537 ± 0.0022	0.6933 ± 0.0032	0.9977 ± 0.0001	0.5945 ± 0.0063	0.9995 ± 0.0001	0.8342 ± 0.0128
GEEL-FI	0.7011 ± 0.0030	0.9578 ± 0.0013	0.6915 ± 0.0029	0.9977 ± 0.0001	0.5790 ± 0.0063	0.9996 ± 0.0001	0.8604 ± 0.0124

ISSN: 1471-2164