Go to ICLR 2025 Conference homepageTri-Comparison Expertise Decision for Drug-Target Interaction Mechanism Prediction
Machine-learned interactions between drugs and human protein targets play a crucial role in efficient and accurate drug discovery. However, the drug-target interaction (DTI) mechanism prediction is actually a multi-class classification problem, which follows a long-tailed class distribution. Existing methods simply address whether interactions can occur and rarely consider the long-tailed DTI mechanism classes. In this paper, we introduce TED-DTI, a novel DTI prediction framework incorporating the divide-and-conquer strategy with tri-comparison options. Specifically, to reduce the learning difficulty of tail classes, we propose an expertise-based divide-and-conquer decision approach that combines the results of multiple independent expertise models for sub-tasks decomposed from the original prediction task. In addition, to enhance the discrimination of similar mechanism classes, we devise a tri-comparison learning strategy that defines the sub-task as the classification of triple options, such as expanding the classification task for classes A and B to include an extra “Neither of them” option. Extensive experiments conducted on various DTI mechanism datasets quantitatively demonstrate the proposed method achieves an approximately 13% performance improvement compared with the other state-of-the-art methods. Moreover, out method exhibits an obvious superiority on the tail classes. Further analysis about the evolvability and generalization of the proposed method reveals the significant potential to be deployed in real-world scenes. Our data and code is included in the Supplementary Materials and will be publicly released after the paper acceptance.