Modelling approaches for chiral chromatography on polysaccharide-based and macrocyclic antibiotic chiral selectors: A review

An overview of molecular modelling approaches, related to chiral separations on polysaccharide-based and macrocyclic antibiotic chiral selectors, is presented. Both atomistic calculations and empirical fitting procedures are discussed. Atomistic calculations, such as docking and molecular dynamics can be used to model the interactions between enantiomers and the chiral stationary phase. This may help obtaining information about the chiral recognition mechanism. Conversely, in empirical fitting procedures, mathematical models for relevant separation parameters are fitted to experimental observations. The latter use theoretical molecular descriptors, calculated from the molecular structure, which are combined into a model to predict a given response, for example, retention. Such relationships, when used in chiral separations, are often called quantitative structure enantioselective retention relationships (QSERR) and an increased interest in them can be observed in the literature. Different regression models are discussed, such as multiple linear regression and partial least squares. (c) 2021 Elsevier B.V. All rights reserved.

Automated summary

This article presents an overview of molecular modelling approaches for chiral separations on polysaccharide-based and macrocyclic antibiotic chiral selectors. Both atomistic calculations and empirical fitting procedures are discussed. Atomistic calculations, such as docking and molecular dynamics, can provide insights into the chiral recognition mechanism by modeling the interactions between enantiomers and the chiral stationary phase. On the other hand, empirical fitting procedures involve mathematical models that are fitted to experimental observations using theoretical molecular descriptors. These models can be used to predict a specific response, such as retention, in chiral separations. The article also discusses different regression models, such as multiple linear regression and partial least squares.