Creation of structure-activity models for predicting characteristics of individual molecules is a mature discipline within the drug discovery industry. There are a great many use cases across the chemical industry where properties are associated with mixtures of chemical substances, e.g. drug formulations, pesticides, cosmetics, hazardous material tracking, and any consumer products with a molecular constituent. A key part of this process is gathering data, which can be done by extracting content from poorly formatted sources (like text) or marking up semi-formatted tables such as spreadsheets. Several key proof of concept examples for creating predictive models using chemical structures, concentrations, and other mixture characteristics will be shown. Methods for working with mixtures data are available both as open source projects and commercial products.

Mixtures QSAR: Modelling collections of chemicals

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.