Discriminating functional and non-functional nuclear-receptor ligands with a conformational selection-inspired machine learning algorithm

• DOI: 10.1016/j.xcrp.2023.101466
• Publication Date: 2023-06-20T14:56:52Z
• Authors: Qinghua Wang; Zhe Wang; Qirui Deng; Sutong Xiang; Rongfan Tang; Yang Yu; Tingjun Hou; Haiping Hao; Huiyong Sun

By targeting the ligand-binding pocket (LBP) and stabilizing corresponding state (agonistic or antagonistic) of a nuclear receptor (NR), functional ligands, namely agonists antagonists, can regulate expression downstream genes. In addition to there exists another category ligands that target LBP an NR, but without functions, which are called non-functional binders (NFBs). practice, it is difficult determine molecular LBP-bound ligand since these molecules bind same position in NR. Therefore, this study, we propose series structure-based machine-learning (ML) models better characterize ligands. Our result shows best-performing model achieves good performance predicting with accuracy >0.85 robust enough be applied real-world tasks, such as identification ligand's out-of-training-set (or external) NRs.