Virtual combinatorial chemical spaces have become valuable in drug discovery. This study examines the composition of the recently published “eXplore” space, which includes around 3.9 trillion virtual product molecules. Several methods, such as FTrees, SpaceLight, and Space-MACS, were used to evaluate the utility of eXplore in retrieving novel chemistry around approved drugs and common Bemis-Murcko scaffolds. A subset of 55 million virtual compounds, representing a commercially attractive and diverse selection of eXplore, was fully enumerated. AI-assisted tools, including ligand-, target-, and fingerprint-based searches, were applied to this subset, revealing new insights into chemical diversity around clinical and lead compounds. Further chemical space analysis, including 2D fingerprints, 3D shape, and electrostatic analyses, demonstrated that eXplore offers a highly diverse chemical space compared to other vendor sets.
Speaker
Peter Maas, Director of Scientific Consulting, eMolecules
Peter E.M. Maas graduated from the Rotterdam University of Applied Sciences in Organic Chemistry. He joined eMolecules (formerly Specs) after completing his internship at Shell Research and Technology Centre, Amsterdam, on the synthesis of ligands for homogeneous catalysis. At eMolecules he specialized in cheminformatics, medicinal chemistry, and early drug discovery. He coordinated numerous research collaborations with universities, pharmaceutical and biotech companies involving chemical and spectral data, solubility, stability for software development, and drug discovery on diverse targets and diseases topics. As Manager Lead Development, he was responsible for all services and expertise within eMolecules to develop lead candidates for drug discovery: library design, medicinal chemistry, formulation studies and patents. This led to the filing of several patents and the establishing of two spin-off companies. In September 2024, he assumed the position of Director of Scientific Consulting.
Register below