Siberian Journal of Life Sciences and Agriculture (Dec 2023)

PREDICTION OF ANTIFUNGAL ACTIVITY OF 1,2,4-TRIAZOLE DERIVATIVES USING QSAR MODELS

  • Alexander L. Osipov

DOI
https://doi.org/10.12731/2658-6649-2023-15-6-1002
Journal volume & issue
Vol. 15, no. 6
pp. 452 – 466

Abstract

Read online

Purpose. Development of QSAR models and investigation of their effectiveness for predicting the antifungal activity of 1,2,4-triazole derivatives against Aspergillus flavus, Aspergillus fumigatus and Trichophyton mentagrophytes. Materials and methods. Experimental data on the antifungal activity of 1,2,4-triazole derivatives against Aspergillus flavus, Aspergillus fumigatus and Trichophyton mentagrophytes were used for scientific research. The data were analyzed using QSAR models based on molecular descriptors automatically generated from structural formulas of 1,2,4-triazole derivatives. Results. New QSAR models for predicting antifungal activity against Aspergillus flavus, Aspergillus fumigatus and Trichophyton mentagrophytes based on six physico-chemical parameters of chemicals (eeig11r, eeig09x, r6m+, belm2, eeig12r) are presented. During the comparative analysis of the developed QSAR models, a model was identified that has the best statistical parameters and does not have multicollinearity: MAE=0.136; MAPE=12.55; forecast accuracy=87.45%; MSE=0.028; RMSE=0.167. Among the six factors, the most significant ones were identified. Conclusion. As a result of the conducted studies, QSAR models for predicting the antifungal activity of 1,2,4-triazole derivatives against Aspergillus flavus, Aspergillus fumigatus and Trichophyton mentagrophytes were identified and analyzed. From one to six molecular descriptors were taken as factors in the models. The factors that make the greatest contribution to the prediction of antifungal activity were evaluated. The best models are selected based on the calculated statistical parameters.

Keywords