De novo design with deep generative models based on 3D similarity scoring

• Published in: Bioorganic & medicinal chemistry Vol. 44; p. 116308
• Main Authors: Papadopoulos, Kostas, Giblin, Kathryn A., Janet, Jon Paul, Patronov, Atanas, Engkvist, Ola
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 15.08.2021
• Subjects: 3D similarity; Deep learning; DRD2; Drug Design; Generative models; Haloperidol - chemical synthesis; Haloperidol - chemistry; Haloperidol - pharmacology; Humans; Ligands; Models, Molecular; Molecular Structure; QSAR; Quantitative Structure-Activity Relationship; Receptors, Dopamine D2 - metabolism; Reinforcement learning; Shape similarity; Structure-Activity Relationship; Deep learning; 3D similarity; QSAR; Reinforcement learning; Generative models; DRD2; Shape similarity
• ISSN: 0968-0896; 1464-3391
• DOI: 10.1016/j.bmc.2021.116308

[Display omitted] We have demonstrated the utility of a 3D shape and pharmacophore similarity scoring component in molecular design with a deep generative model trained with reinforcement learning. Using Dopamine receptor type 2 (DRD2) as an example and its antagonist haloperidol 1 as a starting point in a ligand based design context, we have shown in a retrospective study that a 3D similarity enabled generative model can discover new leads in the absence of any other information. It can be efficiently used for scaffold hopping and generation of novel series. 3D similarity based models were compared against 2D QSAR based, indicating a significant degree of orthogonality of the generated outputs and with the former having a more diverse output. In addition, when the two scoring components are combined together for training of the generative model, it results in more efficient exploration of desirable chemical space compared to the individual components.