In order to accurately predict the covalent protein-ligand binding structure, we developed and validated a hybrid method, dubbed as Cov_DOX, in previous work. Based on a funnel-like strategy, Cov_DOX employs three levels of calculations (molecular mechanics, semi-empirical quantum mechanics, density functional theory) to produce both reliable and affordable predictions of covalent binding structures. Cov_DOX achieved an overall success rate of 81% with RMSD &lt; 2 Å for Top 1 pose prediction in the validation against a test set including 405 crystal structures for covalent protein-ligand complexes, covering various types of the warhead chemistry and receptors. In the failure cases, we found that some of the ligand function group tended to interact with the solution rather than protein cavity. However, all water molecules were removed in the Cov_DOX calculations. Now we consider to put some explicit waters into the cavity to solve this problem. The results show that some of the failure cases affected by solvation are corrected. For example, we improved the RMSD of 3SN8 from 4.4 Å to 0.6 Å. We recommend Cov_DOX be used in the developing scenarios of covalent protein inhibitors and the investigations of the covalent protein-ligand complexes. For a more rigorous forecast, we suggest considering the introduction of explicit water in the energy calculation.

Cov_DOX: A method for structure prediction of covalent protein-ligand bindings

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.