Journal of Pharmacy and Pharmacology
First published: 4 December 2017
Safa Damiati, Norman Smith, Maria Herrero-Zazo, Luigi Martini, Jane Lawrence and David Barlow
Objective: The aim of the work reported here was to use artificial intelligence to model the apparent increase in aqueous solubility of poorly water-soluble drugs afforded by the use of hydrotropes, and thereby to gain a better understanding of the mechanism(s) underlying this phenomenon.
Methods: For the non-steroidal anti-inflammatory drugs (NSAIDs), indomethacin, ibuprofen and mefenamic acid, the solubilising effects of a set of 11 potential hydrotropes were determined experimentally using UV/HPLC. These solubility data, together with 19 molecular descriptors relating to the drugs and hydrotropes, were used to train an ANN.