Table 2 Performance comparison (average ± std) between our model and SOTA methods on unseen drugs, unseen targets, and completely unseen scenarios across the KIBA dataset
Scenario | Method | MSE ↓ | CI↑ | \({r}_{m}^{2}\uparrow\) |
|---|---|---|---|---|
Unseen drug | GraphDTA | 0.471(0.047) | 0.713(0.002) | 0.342(0.007) |
FusionDTA | 0.429(0.031) | 0.748(0.005) | 0.364(0.012) | |
MgraphDTA | 0.425(0.047) | 0.746(0.002) | 0.366(0.016) | |
MSGNN-DTA | 0.426(0.043) | 0.747(0.003) | 0.358(0.022) | |
DMFF(Ours) | 0.408(0.036) | 0.753(0.006) | 0.397(0.020) | |
Unseen target | GraphDTA | 0.469(0.089) | 0.610(0.035) | 0.368(0.057) |
FusionDTA | 0.439(0.062) | 0.685(0.032) | 0.390(0.067) | |
MgraphDTA | 0.435(0.055) | 0.674(0.028) | 0.382(0.047) | |
MSGNN-DTA | 0.438(0.061) | 0.683(0.025) | 0.399(0.054) | |
DMFF(Ours) | 0.410(0.063) | 0.748(0.053) | 0.446(0.064) | |
All unseen | GraphDTA | 0.676(0.113) | 0.601(0.030) | 0.149(0.067) |
FusionDTA | 0.587(0.086) | 0.641(0.023) | 0.193(0.053) | |
MgraphDTA | 0.590(0.094) | 0.626(0.028) | 0.182(0.012) | |
MSGNN-DTA | 0.581(0.079) | 0.648(0.038) | 0.180(0.021) | |
DMFF(Ours) | 0.567(0.082) | 0.667(0.035) | 0.236(0.037) |
