The non-stochastic and stochastic atom-based 3D-chiral quadratic indices were applied to the study of the ?2-adrenoceptor (?2-AR) agonist effect (binding affinities) between a set of 26 stereoisomers of fenoterol, reported with this activity. Linear multiple regression analysis was carried out to predict the ?2-AR binding affinities of the stereoisomers. Two statistically significant QSAR models, able to describe more than the 92% of the variance of the experimental binding affinities, were obtained using non-stochastic (R2 = 0.924 and s = 0.21) and stochastic (R2 = 0.92 and s = 0.22) 3D-chiral linear indices, respectively. The predictability and stability (robustness) of the obtained models (assessed by the leave-one-out cross-validation experiment) yielded values of q2 = 0.893 (scv = 0.237) and q2 = 0.886 (scv = 0.245), respectively. The results obtained with our approach were slightly better than the results of a 3D-QSAR model, obtained with the CoMFA method (R2 = 0.920, q2 = 0.847 and scv = 0.309). The results of our work demonstrate the usefulness of our topological approach for drug discovery of new lead compounds, even in those studies in which the three-dimensional configuration of the chemicals play an important role in the biological activity.

Keywords: non-stochastic and stochastic atom-based 3D-chiral linear index, binding affinity, ?2-adrenoceptor, fenoterol stereoisomer, 3D-QSAR.