NRLiSt BDB, the manually curated nuclear receptors ligands and structures benchmarking database

Area of Science:

• Drug discovery and development
• Structural biology
• Computational chemistry

Background:

• Nuclear receptors (NRs) are critical drug targets.
• Existing databases lack comprehensive, manually curated data for NR drug discovery.
• There is a need for reliable resources to benchmark virtual screening methods for NRs.

Purpose of the Study:

• To create the most exhaustive Nuclear Receptors Ligands and Structures benchmarking database (NRLiSt BDB).
• To provide a manually curated dataset for assessing virtual screening performance against NRs.
• To support the understanding of NR function and the discovery of novel NR-targeting drugs.

Main Methods:

• Compilation of an extensive dataset of 9905 compounds and 339 structures related to NRs.
• Manual review of original literature to ensure data accuracy and identify errors in existing databases like ChEMBL.
• Development of a structured database (NRLiSt BDB) for virtual screening applications.

Main Results:

• The NRLiSt BDB is the most exhaustive NR-focused benchmarking database to date.
• The database contains 9905 compounds and 339 structures, ready for virtual screening.
• Identified and corrected errors in existing datasets through manual curation.

Conclusions:

• The NRLiSt BDB is a powerful, freely available tool for assessing virtual screening methods for NRs.
• The database facilitates a deeper understanding of NR function and modulation.
• NRLiSt BDB will significantly aid in the discovery of new drugs targeting nuclear receptors.