"Identifying a catalyst class to optimize the enantioselectivity of a new reaction, either involving a different combination of known substrate types or an entirely unfamiliar class of compounds, is a formidable challenge. Statistical models trained on a reported set of reactions can help predict out-of-sample transformations 1-5 but often face two challenges: (1) only sparse data are available i.e., limited information on catalyst-substrate interactions, and (2) simple stereoelectronic parameters may fail to describe mechanistically complex transformations."
"6,7 Here we report a descriptor generation strategy that accounts for changes in the enantiodetermining step with catalyst or substrate identity, allowing us to model reactions involving distinct ligand and substrate types. As validating case studies, we collected data on enantioselective nickel-catalyzed C( sp3)-couplings 8 and trained statistical models with features extracted from the transition states and intermediates proposed to be involved in asymmetric induction."
A descriptor generation strategy accounts for changes in the enantiodetermining step with catalyst or substrate identity, enabling modeling of reactions involving distinct ligand and substrate types. Data on enantioselective nickel-catalyzed C(sp3)-couplings were collected and statistical models were trained using features extracted from the transition states and intermediates proposed to govern asymmetric induction. The models predict out-of-sample transformations and permit optimization of poorly performing substrate-scope examples. The models are applicable to unseen ligands and reaction partners. The approach enables quantitative transfer of knowledge from sparse datasets to novel chemical spaces and streamlines catalyst and reaction development.
Read at Nature
Unable to calculate read time
Collection
[
|
...
]