Skip to main content
. 2022 Jun 10;23:224. doi: 10.1186/s12859-022-04763-2

Table 4.

Comparison results on DeepDDI dataset

Method AUROC AUPRC F1
SSP-MLP 92.28 ± 0.18 90.27 ± 0.28 79.71 ± 0.16
GCN 85.53 ± 0.17 83.27 ± 0.31 72.18 ± 0.22
GAT 84.84 ± 0.23 81.14 ± 0.25 73.51 ± 0.38
SEAL-C/AI 92.83 ± 0.19 90.44 ± 0.39 80.70 ± 0.48
MFs 91.54 ± 0.04 89.82 ± 0.24 83.05 ± 0.5
DPDDI 92.79 ± 0.38 91.15 ± 0.52 85.54 ± 0.40
SSI-DDI 96.14 ± 0.06 94.63 ± 0.47 92.27 ± 0.14
DDIMDL 94.85 ± 0.71 93.48 ± 0.07 82.31 ± 0.44
MUFFIN 95.26 ± 0.12 94.47 ± 0.28 91.22 ± 0.48
MIRACLE 95.51 ± 0.27 92.34 ± 0.17 83.60 ± 0.33
MFFGNN 95.39 ± 0.25 96.81 ± 0.16 92.54 ± 0.61

The best results are highlighted in bold