LAUSR

Thiazolidinediones show a large scale of biological activities as a result of ability to perform interactions with a variety of biological targets. Modeling properties of newly synthesized thiazolidinediones is important for both understanding their activity and predicting their interactions. In the paper the chromatographic retention data determined in various RP chromatographic systems (stationary phases RP-CN and RP-18; six aqueous binary mobile phases modified with acetonitrile, methanol, ethanol, propanol, acetone and dioxane) were considered for 13 new 5-arylidene-2,4-thiazolidinediones. In this article, three attempts to find suitable quantitative structure-retention relationship (QSRR) models that quantify retention as a function ofmolecular descriptors had been presented. Models built for RP-18 show generally better multiple R but are also mostly monoparametric with logP as the dominant descriptor. More informative from the standpoint of molecular interactions are QSRR models for RP-CN. The quality of those models depends of the mobile phase modifier (the best was obtained for acetone and the worst for propanol as modifier). Since all QSRR models use extrapolated retention as a property which is indirectly connected with plasma protein binding further assessment of plasma protein binding should be based on extrapolated retention on a RP-CN stationary phase instead of standard RP-18.