Prediction of Drug Intestinal Absorption by New Linear and Non-Linear QSPR

ALAN TALEVI; M. GOODARZI, E. ORTIZ, P.R. DUCHOWICZ, G. PESCE, E.A. CASTRO, ; PESCE GUIDO; EDUARDO A. CASTRO,; BRUNO BLANCH, LUIS

EUROPEAN JOURNAL OF MEDICAL CHEMISTRY

ELSEVIER FRANCE-EDITIONS SCIENTIFIQUES MEDICALES ELSEVIER

Lugar: Paris; Año: 2011 vol. 46 p. 218 - 228

0223-5234

In order to minimize the high attrition rate that usually characterizes drug research and development projects, current medicinal chemists aim to characterize both pharmacological and ADME profiles at the beginning of drug R&D initiatives. Thus, the development of ADME High-Throughput Screening in vitro and in silico ADME models has become an important growing research area. Here we present new linear and non-linear predictive QSPR models to predict the human intestinal absorption rate,which are derived from a medium sized, balanced and diverse training set of organic compounds. The structureeproperty relationships so obtained involve only 4molecular descriptors, and display an excellent ratio ofnumber of cases to number of descriptors. Their adjustment of the training set data together with the performance achieved during the internal and external validation procedures are comparable to previously reported modeling efforts