The authors proposed a multi-task deep learning model, to accurately predict drug-target affinity and generate target-aware drugs. In the proposed model, the authors developed the FetterGrad algorithm to mitigate gradient conflicts between both tasks. Through comprehensive experimentation, the authors have shown, that their proposed model effectively predicts affinity and generates target-aware drugs thus enhancing the drug discovery process.
- Pir Masoom Shah
- Huimin Zhu
- Min Li
