Skip to main content

Table 3 fivefold cross-validation results performed by the three models GRLMN, GRLMN_A, and GRLMN_M

From: An effective drug-disease associations prediction model based on graphic representation learning over multi-biomolecular network

Model ACC (%) SEN (%) SPE (%) PRE (%) MCC (%) AUC (%)
GRLMN 79.84 ± 0.50 80.03 ± 0.95 79.64 ± 0.22 79.72 ± 0.28 59.68 ± 1.00 87.90 ± 0.54
GRLMN_A 73.91 ± 0.32 75.43 ± 0.77 72.39 ± 0.34 73.20 ± 0.20 47.84 ± 0.64 81.12 ± 0.28
GRLMN_M 77.58 ± 0.54 78.41 ± 1.07 76.75 ± 0.46 77.14 ± 0.39 55.18 ± 1.10 85.61 ± 0.42