In Silico Analysis of USP7 Inhibitors Based on Building QSAR Models and Fragment Design for Screening Marine Compound Libraries

Abstract

USP7 is highly expressed in a variety of tumors and is thought to play a major role in cancer development. However, there are no drugs available to target USP7, so there is a need to develop new USP7 inhibitors. In this study, AutoQSAR, multiple linear regression, and Naive Bayesian models were constructed using 543 compounds and used to analyze marine compounds. After selecting 240 small molecules for molecular docking with Maestro, MOE, and GOLD, better small molecules than the positive compound P217564 were screened. The molecular structure of “1, 2-dibromobenzene” was optimized to improve the binding effect of the protein, and 10 optimized compounds in ADMET performed well during the screening process. To study the dynamic combination of protein–ligand effect consistency with static molecular docking, 100ns molecular dynamics simulations of candidate compound 1008-1, reference compound P217564, and negative-positive GNE2917 were conducted. The results of molecular docking and molecular dynamics simulation analysis showed that compound 1008-1 maintained a stable conformation with the target protein. Thus, the comprehensive analysis suggests that compound 1008-1 could provide new possibilities for USP7 covalent inhibitor candidates.