CDPMF-DDA: contrastive deep probabilistic matrix factorization for drug-disease association prediction

BMC Bioinformatics

Table 1 The results of 10 times of tenfold cross-validation

Datasets	LBMFF	SCPMFDD	SCPMF	MKGCN	MNGACDA	CDPMF-DDA
AUROCs
Fdataset	0.7953 ± 0.035	0.7740 ± 0.001	0.8957 ± 0.001	0.8870 ± 0.001	0.8179 ± 0.005	0.9477 ± 0.003
Cdataset	0.9069 ± 0.001	0.7937 ± 0.001	0.9117 ± 0.002	0.9109 ± 0.001	0.8406 ± 0.005	0.9548 ± 0.005
LRSSL	0.9139 ± 0.002	0.7668 ± 0.001	0.8977 ± 0.001	0.8596 ± 0.001	0.7936 ± 0.002	0.9401 ± 0.005
Avg	0.8720	0.7782	0.9017	0.8858	0.8173	0.9475
AUPRs
Fdataset	0.0266 ± 0.012	0.0056 ± 0.000	0.3451 ± 0.006	0.4471 ± 0.001	0.2494 ± 0.001	0.5540 ± 0.005
Cdataset	0.2176 ± 0.006	0.0055 ± 0.000	0.4140 ± 0.004	0.5647 ± 0.001	0.3396 ± 0.001	0.5967 ± 0.005
LRSSL	0.1740 ± 0.005	0.0038 ± 0.000	0.2710 ± 0.002	0.3518 ± 0.001	0.1799 ± 0.001	0.3520 ± 0.005
Avg	0.1394	0.00498	0.3433	0.4545	0.2563	0.5009

a ± b indicates that a fluctuates within the range of b. bold values represent the second-best results across the three datasets

ISSN: 1471-2105