论文部分内容阅读
代替原子序数,采用有意义的属性,例如基团尺寸,来比较手性中心键连4个基团的大小,并通过各基团属性的排序判断手性中心的构型为类R/S。以所得到的新构型为基础,扩展了基于分子结构和距离矩阵的Am指数,提出了新的手性Am指数,以其与物化手性描述符相结合来表征一个分子的手性,并应用于脂化酶作为催化剂条件下,手性伯醇产物的立体选择预测以手性指数和Random Forests建立模型,通过交叉验证,能正确预测整个数据集的90%的对映异构体。
Instead of atomic numbers, the size of chiral center bond with four groups is compared with meaningful attributes such as group size, and the chiral center configuration is judged as R / S based on the order of each group’s properties. Based on the new configuration obtained, we extend the Am index based on the molecular structure and distance matrix and propose a new chiral Am index, which can be used to characterize the chirality of a molecule by combining it with the materialized chiral descriptor Applying Lipase as Catalyst, the stereoselectivity of chiral primary alcohol products was predicted using a chiral index and Random Forests model. By cross-validation, 90% of the enantiomer of the entire data set can be correctly predicted.