Prediction of in vitro metabolic stability of calcitriol analogs by QSAR.

Abstract

The metabolic stability of a drug is an important property for potential drug candidates. Measuring this property, however, can be costly and time-consuming. The use of quantitative structure-activity relationships (QSAR) to estimate the in vitro stability is an attractive alternative to experimental measurements. A data set of 130 calcitriol analogs with known values of in vitro metabolic stability was used to develop QSAR models. The analogs were encoded with molecular structure descriptors computed mainly with the commercial software QikProp and DiverseSolutions. Variable selection was carried out by five different variable selection techniques and Partial Least Squares Regression (PLS) models were generated from the 130 analogs. The models were used for prediction of the metabolic stability of 244 virtual calcitriol analogs. Twenty of the 244 analogs were selected and the in vitro metabolic stability was determined experimentally. The PLS models were able to predict the correct metabolic stability for 17 of the 20 selected analogs, corresponding to a prediction performance of 85%. The results clearly demonstrate the utility of QSAR models in predicting the in vitro metabolic stability of calcitriol analogs.