DockThor-VS: A Free Platform for Receptor-Ligand Virtual Screening

• Published in: Journal of molecular biology 2024-09, Vol.436 (17), p.168548, Article 168548
• Main Authors: Guedes, Isabella Alvim, Pereira da Silva, Matheus Müller, Galheigo, Marcelo, Krempser, Eduardo, de Magalhães, Camila Silva, Correa Barbosa, Helio José, Dardenne, Laurent Emmanuel
• Format: Article
• Language: English
• Subjects: Algorithms; amino acids; COVID-19 infection; data collection; drug discovery; Drug Discovery - methods; drugs; Humans; Internet; Ligands; molecular biology; molecular docking; Molecular Docking Simulation; peptides; prediction; Protein Binding; Proteins - chemistry; Proteins - metabolism; protonation; Software; structure-based drug design; user interface; User-Computer Interface; virtual screening
• ISSN: 0022-2836; 1089-8638; 1089-8638
• DOI: 10.1016/j.jmb.2024.168548

[Display omitted] •The DockThor-VS web server is a free platform for virtual screening for the scientific community that has been under continuous development since its first release in 2013.•In 2023, more than 27,000 jobs were submitted to the DockThor-VS web server, and we could track more than 330 citations to three of our methodological papers (published since 2020).•The DockThor suite has accuracy compared to the state-of-the-art protein–ligand docking methods for pose and affinity predictions.•We provided the ready-to-dock structures of relevant COVID-19-related structures, including clinically relevant non-synonym variations and distinct conformations representative of the flexibility of each receptor.•Recently, we added new COVID-19-related curated targets from SARS-CoV-2 (Nsp13, and Nsp3 - macro domain) and human host (TMPRSS2 and cathepsin B)•Recently, we included an updated database of FDA-approved drugs with protonation and tautomeric states within different pH ranges ready to be used in virtual screening experiments,•Recently, we include new datasets of compounds from natural sources ready to be used in virtual screening experiments. The DockThor-VS platform (https://dockthor.lncc.br/v2/) is a free protein–ligand docking server conceptualized to facilitate and assist drug discovery projects to perform docking-based virtual screening experiments accurately and using high-performance computing. The DockThor docking engine is a grid-based method designed for flexible-ligand and rigid-receptor docking. It employs a multiple-solution genetic algorithm and the MMFF94S molecular force field scoring function for pose prediction. This engine was engineered to handle highly flexible ligands, such as peptides. Affinity prediction and ranking of protein–ligand complexes are performed with the linear empirical scoring function DockTScore. The main steps of the ligand and protein preparation are available on the DockThor Portal, making it possible to change the protonation states of the amino acid residues, and include cofactors as rigid entities. The user can also customize and visualize the main parameters of the grid box. The results of docking experiments are automatically clustered and ordered, providing users with a diverse array of meaningful binding modes. The platform DockThor-VS offers a user-friendly interface and powerful algorithms, enabling researchers to conduct virtual screening experiments efficiently and accurately. The DockThor Portal uti