Identifying general reaction conditions by bandit optimization.

Autor: Wang JY; Department of Chemistry, Princeton University, Princeton, NJ, USA.; Department of Chemistry and Biochemistry, University of California, Los Angeles, CA, USA., Stevens JM; Chemical Process Development, Bristol Myers Squibb, Summit, NJ, USA., Kariofillis SK; Department of Chemistry, Princeton University, Princeton, NJ, USA.; Department of Chemistry and Biochemistry, University of California, Los Angeles, CA, USA.; Department of Chemistry, Columbia University, New York, NY, USA., Tom MJ; Department of Chemistry and Biochemistry, University of California, Los Angeles, CA, USA., Golden DL; Chemical Process Development, Bristol Myers Squibb, Summit, NJ, USA., Li J; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., Tabora JE; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., Parasram M; Department of Chemistry, Princeton University, Princeton, NJ, USA.; Department of Chemistry, New York University, New York, NY, USA., Shields BJ; Department of Chemistry, Princeton University, Princeton, NJ, USA.; Molecular Structure and Design, Bristol Myers Squibb, Cambridge, MA, USA., Primer DN; Chemical Process Development, Bristol Myers Squibb, Summit, NJ, USA.; Loxo Oncology at Lilly, Louisville, CO, USA., Hao B; Janssen Research and Development, Spring House, PA, USA., Del Valle D; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., DiSomma S; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., Furman A; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., Zipp GG; Discovery Synthesis, Bristol Myers Squibb, Princeton, NJ, USA., Melnikov S; Spectrix Analytical Services, North Haven, CT, USA., Paulson J; Chemical Process Development, Bristol Myers Squibb, New Brunswick, NJ, USA., Doyle AG; Department of Chemistry, Princeton University, Princeton, NJ, USA. agdoyle@chem.ucla.edu.; Department of Chemistry and Biochemistry, University of California, Los Angeles, CA, USA. agdoyle@chem.ucla.edu.
Jazyk: angličtina
Zdroj: Nature [Nature] 2024 Feb; Vol. 626 (8001), pp. 1025-1033. Date of Electronic Publication: 2024 Feb 28.
DOI: 10.1038/s41586-024-07021-y
Abstrakt: Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries 1-6 . Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models 7-10 to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.
(© 2024. The Author(s), under exclusive licence to Springer Nature Limited.)
Databáze: MEDLINE
Externí odkaz: