Skip to main content
. 2022 Nov 23;39(1):btac731. doi: 10.1093/bioinformatics/btac731

Table 5.

Comparison of Perceiver CPI and other SOTA competitors on novel-hard pair setting

Model MSE CI
DeepDTA (Öztürk et al., 2018) 0.948(±0.218) 0.565(±0.040)
DeepConvDTI (Lee et al., 2019) 0.768(±0.290) 0.571(±0.052)
TransformerCPI (Chen et al., 2020) 0.806(±0.254) 0.508(±0.071)
GraphDTA (GINs) (Nguyen et al., 2021) 0.931(±0.314) 0.542(±0.070)
HyperattentionDTI (Zhao et al., 2022) 0.873(±0.246) 0.600(±0.049)
Perceiver CPI (ours) 0.701(±0.244) 0.609(±0.072)