RESUMO
Structural diversification of lead molecules is a key component of drug discovery to explore chemical space. Late-stage functionalizations (LSFs) are versatile methodologies capable of installing functional handles on richly decorated intermediates to deliver numerous diverse products in a single reaction. Predicting the regioselectivity of LSF is still an open challenge in the field. Numerous efforts from chemoinformatics and machine learning (ML) groups have made strides in this area. However, it is arduous to isolate and characterize the multitude of LSF products generated, limiting available data and hindering pure ML approaches. We report the development of an approach that combines a message passing neural network and 13C NMR-based transfer learning to predict the atom-wise probabilities of functionalization for Minisci and P450-based functionalizations. We validated our model both retrospectively and with a series of prospective experiments, showing that it accurately predicts the outcomes of Minisci-type and P450 transformations and outperforms the well-established Fukui-based reactivity indices and other machine learning reactivity-based algorithms.
Assuntos
Descoberta de Drogas , Redes Neurais de Computação , Estudos Prospectivos , Estudos Retrospectivos , Descoberta de Drogas/métodos , Aprendizado de MáquinaRESUMO
An experimental approach is described for late-stage lead diversification of frontrunner drug candidates using nanomole-scale amounts of lead compounds for structure-activity relationship development. The process utilizes C-H bond activation methods to explore chemical space by transforming candidates into newly functionalized leads. A key to success is the utilization of microcryoprobe nuclear magnetic resonance (NMR) spectroscopy, which permits the use of low amounts of lead compounds (1-5 µmol). The approach delivers multiple analogues from a single lead at nanomole-scale amounts as DMSO-d6 stock solutions with a known structure and concentration for in vitro pharmacology and absorption, distribution, metabolism, and excretion testing. To demonstrate the feasibility of this approach, we have used the antihistamine agent loratadine (1). Twenty-six analogues of loratadine were isolated and fully characterized by NMR. Informative SAR analogues were identified, which display potent affinity for the human histamine H1 receptor and improved metabolic stability.