DINC-COVID: A webserver for ensemble docking with flexible SARS-CoV-2 proteins

Area of Science:

• Computational Biology
• Drug Discovery
• Structural Biology

Background:

• Extensive research on SARS-CoV-2 proteins has yielded numerous structures and virtual screening efforts.
• Previous computational drug discovery initiatives have shown limited success in identifying effective SARS-CoV-2 inhibitors.
• Receptor flexibility is an often overlooked factor in computational drug discovery for SARS-CoV-2.

Purpose of the Study:

• To implement a computational tool for ensemble docking that accounts for SARS-CoV-2 protein flexibility.
• To provide a user-friendly webserver (DINC-COVID) with pre-computed protein ensembles for docking studies.
• To validate the DINC-COVID platform for identifying potential SARS-CoV-2 inhibitors.

Main Methods:

• Extracted representative protein ensembles from the Protein Data Bank and molecular dynamics simulations.
• Developed a webserver, DINC-COVID, offering twelve pre-computed SARS-CoV-2 protein ensembles for ensemble docking.
• Validated the platform using experimental data on inhibitors, correlating docking-derived binding energies with experimental affinities.

Main Results:

• DINC-COVID demonstrated good correlations between predicted binding energies and experimentally determined affinities.
• The platform successfully identified potential inhibitors, particularly with large ligands captured in alternative receptor conformations.
• Ensembles within DINC-COVID capture diverse ranges of receptor flexibility, aiding in the discovery of alternative ligand binding modes.

Conclusions:

• Accounting for receptor flexibility is crucial for successful computational drug docking studies.
• The DINC-COVID webserver provides a valuable platform for identifying novel inhibitors against SARS-CoV-2 proteins.
• This approach enhances drug discovery efforts by incorporating dynamic protein structures.