gGATLDA: lncRNA-disease association prediction based on graph-level graph attention network

BMC Bioinformatics

Table 3 Experiment results of six methods on Dataset3 under CVP setting

	gGATLDA	BiWalkLDA	SIMCLDA	MFLDA	BiGAN	GCRFLDA
AUC	0.9888 ± 0.0065	0.8185 ± 0.0024	0.8465 ± 0.0030	0.8478 ± 0.0048	0.9045 ± 0.0185	0.9583 \(\pm\) 0.0055
Precision	0.7980 ± 0.0367	0.6370 ± 0.0033	0.7247 ± 0.0142	0.8667 ± 0.1310	0.6572 ± 0.0073	0.9020 \(\pm\) 0.0052
Recall	0.9913 ± 0.0078	0.7297 ± 0.0121	0.8475 ± 0.0162	0.6942 ± 0.1089	0.9495 ± 0.0132	0.8632 \(\pm\) 00,202
AUPR	0.9890 ± 0.0060	0.8416 ± 0.0031	0.8450 ± 0.0053	0.8860 ± 0.0032	0.9058 ± 0.0192	0.9548 \(\pm\) 0.0090
Accuracy	0.8670 ± 0.0271	0.6568 ± 0.0032	0.7623 ± 0.0065	0.7652 ± 0.0867	0.7270 ± 0.0088	0.9103 \(\pm\) 0.0044
F1-Score	0.8830 ± 0.0217	0.6801 ± 0.0049	0.7810 ± 0.0022	0.7523 ± 0.0324	0.7767 ± 0.0068	0.8817 \(\pm\) 0.0130

ISSN: 1471-2105