DrugSpaceX: a large screenable and synthetically tractable database extending drug space

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Exploring vast chemical space for novel drug candidates is a key challenge in drug discovery.
• Synthesizable and structurally novel compounds with desired biological properties are sought after.

Purpose of the Study:

• To introduce DrugSpaceX, a database designed for efficient exploration of chemical space.
• To provide a resource for virtual screening and ligand identification in drug discovery.

Main Methods:

• Developed DrugSpaceX based on expert-defined transformations of approved drug molecules.
• Generated over 100 million transformed chemical products for virtual screening.
• Included subsets for download (diversity, drug-like, lead-like, fragment-like) for ligand identification and optimization.

Main Results:

• DrugSpaceX offers >100 million transformed molecules with high structural novelty, diversity, and 3D coverage.
• Demonstrated practical application through a case study on discoidin domain receptor 1 (DDR1) for hit compound identification.
• Provides transformation location data to aid medicinal chemists in strategy planning.

Conclusions:

• DrugSpaceX is a valuable resource for accelerating drug discovery by enabling efficient virtual screening.
• The database facilitates the identification and optimization of ligands for therapeutic targets.
• Its features support medicinal chemists in designing novel drug candidates.