The derivation of a chiral substituent code for secondary alcohols and its application to the prediction of enantioselectivity

A chiral substituent code was proposed based on the features of secondary alcohols, in which a chiral center is attached to two substituents in addition to OH and H substituents. The new chirality code, which was generated by predefining positional information of four substituents attached to stereocenter, was applied to two datasets composed of secondary alcohols as the enantioselective products of asymmetric reactions. In the first dataset, the chemical reaction was catalyzed by a biocatalyst, lipase from Candida rugosa. The catalyst for the second dataset was (−)-diisopinocampheylchloroborane. The structure–enantioselectivity relationship models were constructed using random forests with the chiral substituent code as the input. The resulting models were assessed both in terms of single enantiomers and pairs of enantiomers. Satisfactory results were obtained for both datasets. Although the chiral substituent code was specifically developed for secondary alcohols, it can easily be extended to represent chiral compounds possessing a specific chiral center bonded to two variable substituents.