Abstract
A QSAR model has been developed for 1,3-dimethylxanthines as adenosine receptor antagonists. The model is capable of predicting the affinity toward both the A1 and A2 receptors. Constitutional, geometrical, topological, electronic descriptors (computed at the ab initio 6-31G level), and some empirical descriptors related to the hypophilicity were computed and analyzed. A two step computational strategy was adopted to select the descriptors relevant to the A1 or the A2 affinity. In the first step, each of the four main groups of descriptors is treated independently. Multiple regression analysis lead to a set of equations that reflect the weight of each of the studied descriptors. The most relevant of these descriptors were grouped, and a new multiple regression analysis has been carried out and arrived at the final QSAR model. These QSAR equations account for almost all the A2 and an appreciable part of the A1 affinity. The proposed model has been examined as a general tool of predicting the activity toward the adenosine receptor sites. A validation set of 22 xanthines were selected, and their activities were computed using the proposed QSAR model. The correspondence between the predicted and observed activities is excellent. Anova statistical analysis on the data of the validation set elaborates on the quality of these fits.