Skip to main content

Table 10 Comparison results of different PPIs prediction methods on S.cerevisiae (core subset)

From: SDNN-PPI: self-attention with deep neural network effect on protein-protein interaction prediction

Methods ACC(%) Spec(%) Sens(%) Prec(%) MCC(%) AUC(%)
DeepPPI[34] 94.43 ± 0.30 N/A 92.06 ± 0.36 96.65 ± 0.59 88.97 ± 0.62 N/A
DeepFE-PPI[39] 94.78 ± 0.61 N/A 92.99 ± 0.66 96.45 ± 0.87 89.62 ± 1.23 N/A
LightGBM-PPI[27] 95.07 97.94 92.21 97.82 90.30 N/A
Bio2Vec[59] 93.30 N/A 92.70 93.55 87.49 97.20
StackPPI[50] 94.64 96.46 92.81 96.33 89.34 N/A
GTB-PPI[28] 95.15 ± 0.25 N/A 92.21 ± 0.36 97.97 ± 0.60 90.45 ± 0.53 N/A
AE-LGBM[60] 95.40 ± 0.20 98.70 ± 0.20 92.10 ± 0.30 N/A 91.00 ± 0.40 N/A
GcForest-PPI[41] 95.44 ± 0.18 N/A 92.72 ± 0.44 98.05 ± 0.25 91.02 ± 0.35 N/A
SDNN-PPI 95.48 ± 0.37 97.23 ± 0.66 93.80 ± 0.39 97.13 ± 0.66 91.02 ± 0.74 98.63