Molecular simulation approach integrated with deep learning model for the discovery of Tankyrase inhibitors

Abstract

Tankyrases belong to the poly (ADP-ribose) polymerase family epitomized as a novel group of medicinal targets with various prospective diseased conditions and it is appraised to be a challenging drug target for the intervention of multiple cancers. Thus, the principal objective of our study is to explore the dual-site selective tankyrase 1 inhibitor by employing the pharmacophore strategy. Initially, the ligand-based pharmacophore study generated five featured pharmacophore hypothesis, which was then employed for the database screening. The screened molecules were scrutinized through docking, MM/GBSA calculations and molecular simulation alongside pharmacokinetics properties. The analysis yielded potent dual-site tankyrase 1 inhibitors such as nebivolol and ondansetron from the DrugBank repository. Notably, the recognized lead molecules were perceived to have higher XP GScore and binding energy scores. Subsequently, simulation studies were executed to validate the structural stability of the lead molecules. It is worth mentioning that the existence of benzopyran and carbazole scaffolds in the lead molecules displayed anti-neoplastic activity and also facilitate the effective binding with tankyrase 1 protein. Ultimately, the IC50 values of the lead molecules were examined against the NCI-H596 cell line using a deep learning model. Indeed, these results are of immense importance and provide a clue to the experimental biologist in developing a potent tankyrase 1 dual-site inhibitor.