Prediction of antihistamine activity according to QSAR methods

Antihistamines are responsible for blocking histamine receptors, thus reducing the effects of this amine on the body. However, the experimental classification of these compounds is accompanied by several limitations such as the high time invested and the consumption of large amounts of resources. QSAR methods reduce the cost and time spent discovering new drugs. The objective of the present study was to model the antihistamine activity of a series of compounds reported in the literature for the identification of new therapeutic candidates. For this, the calculation of the spectral moments of the adjacency matrix between edges of the molecular graph with different parameters that characterize the molecules of 90 active and 250 inactive compounds was carried out. using the MODESLAB methodology. 91 descriptors related to the activity of these drugs were calculated, which were used in a training series divided into two groups. In order to identify the descriptors that best discriminate and define their set of functions, a linear discriminant analysis was developed using the step-bystep inclusion method using the IBM SPSS version 22 statistical software. A function was generated that constitutes linear combinations of eight molecular descriptors, which encode both steric and electronic information of the molecules of each group. The functions obtained present a Wilks Lambda of (0.421) and a high canonical correlation of (0.8351), which shows its discriminating power, and it makes it possible to use the descriptors included in them in future studies of the structure-property or structure-activity relationship. The results obtained suggest the use of this model with a high predictive value (prediction coefficient of 82.19%) in the determination of compounds with antihistamine activity.