Versional Improvements in LogS Prediction

At ACD/Labs we continue to improve the accuracy of our physicochemical predictions with each release. By increasing the size and structural diversity of the internal database and with implementation of enhanced algorithms, we have continued to deliver, release to release, significant improvements in the quality of our predictions.

In version 11.0, improvements to solubility predictions are a result of significant enhancements to the underlying properties of logP and pK_a. For more details read "What's New in ACD/Solubility DB".

Version-to-version improvements to the prediction accuracy of solubility have been followed for a number of years using a small dataset of 119 compounds (pesticides). The figure below illustrates changes in prediction accuracy from version 5 to the current version.

By version 11.0, compound solubility is predicted within 1.0 log unit for 89% of the dataset. Enhancement from version 9.0 to 11.0 is most noticeable in the error range 0.0-0.5 log units. In version 11.0, solubility for 65% of the compounds is predicted within 0.5 log units compared to 61% in version 9.0.

This comparison of data is by no means a comprehensive study of solubility and accuracy of prediction; rather it is an internal benchmark for ACD/Labs to evaluate the evolution of the product, and a study of the enhancement of the algorithm. Our clients have ranging accuracy needs based on their applications, and are always encouraged to test their data when contemplating deployment options.