Rational design and synthesis of pyrazole derivatives as potential SARS-CoV-2 Mpro inhibitors: An integrated approach merging combinatorial chemistry, molecular docking, and deep learning

Abstract

The global impact of SARS-CoV-2 has highlighted the urgent need for novel antiviral therapies. This study integrates combinatorial chemistry, molecular docking, and deep learning to design, evaluate and synthesize new pyrazole derivatives as potential inhibitors of the SARS-CoV-2 main protease (M^pro). A library of over 60,000 pyrazole-based structures was generated through scaffold decoration to enhance chemical diversity. Virtual screening employed molecular docking (ChemPLP scoring) and deep learning (DeepPurpose), with consensus ranking to identify top candidates. Binding free energy calculations refined the selection, revealing critical structural features such as tryptamine and N-phenyl fragments for M^pro binding. High-temperature solvent-free amidation allowed the synthesis of a selected derivative. Final compounds demonstrated favorable drug-likeness properties based on Lipinski’s and Veber’s rules. This work highlights the integration of computational and synthetic strategies to accelerate the discovery of M^pro inhibitors and provides a framework for future antiviral development.