ZUPANS DESCRIPTORS IN QSAR APPLIED TO THE STUDY OF A NEW CLASS OF CARDIOTONIC AGENTS

Recently a new class of molecular descriptors has been proposed and us ed in QSAR with simulated data and with regression performed by neural networks. In the present paper these descriptors (Zups, from the name of their author, Juri Zupan) have been slightly modified and then app lied to a real data set with the aim of studying the structure-activit y relationships of a new class of cardiotonics. Forty-one molecules (t hirty-seven milrinone analogues, the two lead compounds amrinone and m ilrinone, and two commercial products) have been studied using classic al chemometrical techniques such as PCA (Principal Components Analysis ) and PLS (Partial Least Squares regression). Zups describe essentiall y the local geometry of the molecules. They show promising performance s, as compared with other classical geometrical descriptors (as molecu lar volume, etc.), both in that regards the overall performances, meas ured by the C.V. Explained variance and in the interpretability of the regression equation. However they have not all the requirements of a good structure representation. Moreover some selectable parameters see m to have a great importance, so that the refinement of the regression model requires time and the evaluation step must be performed in cond ition of full-validation, because predictive optimisation is used in t he selection of parameters, and the final model must be checked on mol ecules never used to refine the model or, in this case, the parameters of the structure representation.