![Author ORCID: We display the ORCID iD icon alongside authors names on our website to acknowledge that the ORCiD has been authenticated when entered by the user. To view the users ORCiD record click the icon. [opens in a new tab]](https://chemrxiv.org/engage/assets/public/chemrxiv/images/logos/orcid.png)
Abstract
Since known approved drugs like liponavir and ritonavir failed to cure SARS-CoV-2
infected patients, it is high time to generate new chemical entities against this virus.
3CL main protease acts as key enzyme for the growth of a virus which acts as
biocatalyst and helps to break protein and ultimately in the growth of coronavirus.
Based on a recently solved structure (PDB ID: 6LU7), we developed a novel
advanced deep Q-learning network with the fragment-based drug design
(ADQN-FBDD) along with variational autoencoder with KL annealing and circular
annealing for generating potential lead compounds targeting SARS-CoV-2 3CLpro.
Structure-based optimization policy (SBOP) is used in reinforcement learning. The
reason for using variational autoencoders is that it does not deviate much from actual
inhibitors, but since VAE suffers from KL diminishing we have used KL annealing
and circular annealing to address this issue. Researchers can use this compound as
potential drugs against SARS-CoV-2
