New COVID-19 protease inhibitors

designed by AI with recurrent neural networks, molecular docking, and QSAR

The project aims at finding new effecive inhibitors for COVID-19 protease

The generator is based on our previous work on MDMX inhibitors, published in Journal of Coputer-Aided Molecular Design: Zhonghua Xia, Pavel Karpov, Grzegorz Popowicz and Igor V. Tetko. Focused Library Generator: case of Mdmx inhibitors. Description of the method available here

The data on the site is refreshing in real-time. When the AI generates a new perspective structure, the system looks the compound in databases of available-on-demand chemicals and adds it to the respective list (Statistics panel on the right) if records were found. Currently, the following databases are supported: ZINC,Enamine Real Database, and ChEMBL. Due to the dynamic nature of the overall system, the resulting Top list may be changed because of finding more promising structures. The system can be easily adapted to other targets, thus, supporting drug-development projects.


Download

id SMILES Docking pIC50 (SARS) pIC50 (HIV) logS
91262 Cc1ccncc1C(C)NC(=O)C(Cc1c[nH]c2ccccc12)NC(=O)c1ccc(F)cc1 - 5.83 6.44 -4.5
91276 Cc1cnc2c(S(=O)(=O)NC(Cc3ccccc3)C(=O)Nc3ccccc3)cccc2c1 - 5.56 6.42 -4.34

Generator summary

Random prospective compound




O=C(N[C@@H](Cc1nnnc2ccccc12)C(=O)Nc1cccs1)c1cccc2nnnnc12

COVID-19 protease

The structure has been recently solved, PDB code 6lu7. 6ul7 structure
Developed in group of Cheminformatics in the Institute of Structural Biology at Helmholtz-Zentrum Muenchen, Tetko group
With support of BIGCHEM GmBH (bigchem.de) and RULIS Ltd. (rulis.ru).