The distribution coefficient or LogD is a commonly used measure of the lipophilicity of a small molecule. The computed LogD is a key parameter used by the Lipinski Rule-of-5 and its Agchem equivalent, the Tice rules. These rules increase the likelihood of successful products. Several commercially available software tools provide models that calculate LogD values based on the chemical structure of organic molecules. Unfortunately, the applicability domain of such models is often restricted to primarily pharmaceuticals, and thus, do not provide very accurate results for compounds of interest in crop protection. The present work describes the development and implementation of machine learning models that enhance calculated vs. measured distribution coefficients (LogD) of small agrochemicals utilizing internally measured data at various pH values. The errors of the composite models are close to the experimental error.

Improved lipophilicity (clogD) QSAR models for agrochemicals

2019 Fall National Meeting

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