GalaxySagittarius-AF: Predicting Targets for Drug-Like Compounds in the Extended Human 3D Proteome.

[Display omitted] • Predicting interactions between a chemical compound and a biological system is crucial for understanding the physiological effects of the compound and for drug discovery purposes. • The web server GalaxySagittarius-AF provides structure-based predictions for protein targets of a given druglike compound in the human 3D proteome, along with their complex structures with the predicted protein targets. • Compared to the previous version, GalaxySagittarius-AF has expanded its potential protein targets from 6,037 to 15,956 unique human proteins, utilizing protein structures predicted by AlphaFold. • Furthermore, the web server has been significantly enhanced, now operating two to three times faster and offering comprehensive information about the predicted targets. In recent years, advancements in deep learning techniques have significantly expanded the structural coverage of the human proteome. GalaxySagittarius-AF translates these achievements in structure prediction into target prediction for druglike compounds by incorporating predicted structures. This web server searches the database of human protein structures using both similarity- and structure-based approaches, suggesting potential targets for a given druglike compound. In comparison to its predecessor, GalaxySagittarius, GalaxySagittarius-AF utilizes an enlarged structure database, incorporating curated AlphaFold model structures alongside their binding sites and ligands, predicted using an updated version of GalaxySite. GalaxySagittarius-AF covers a large human protein space compared to many other available computational target screening methods. The structure-based prediction method enhances the use of expanded structural information, differentiating it from other target prediction servers that rely on ligand-based methods. Additionally, the web server has undergone enhancements, operating two to three times faster than its predecessor. The updated report page provides comprehensive information on the sequence and structure of the predicted protein targets. GalaxySagittarius-AF is accessible at https://galaxy.seoklab.org/sagittarius%5faf without the need for registration. [ABSTRACT FROM AUTHOR]