SUPPORT VECTOR MACHINE MODEL FOR PREDICTING ACTIVITY OF INHIBITORS AGAINST SARS-COV 3CLpro ENZYME

ABSTRACT

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a beta coronavirus which led to coronavirus disease-2019 (COVID-19) and has threatened global public health and economy. Currently there is no specific medicine for COVID-19. So there is an urgent need to develop broad-spectrum anti-coronavirus drugs. The SARS-CoV 3-chymotrypsin-like protease (3CLpro) is highly conservative in beta-coronavirus and becomes viable target used for anti-SARS drugs. Support vector machine (SVM) algorithm was used to build quantitative structure–activity relationships (QSARs) for the activity (logIC50) of 204 inhibitors for SARS-CoV 3CLpro enzyme. Seven molecular descriptors were selected for the optimal SVM model with parameters C = 250 and γ = 0.15, which has root-mean-square (rms) errors being 0.341 (training set), 0.337 (validation set) and 0.336 (test set). Comparison with other models in the literature shows that the SVM model was proved to be satisfactory although the SVM model in this paper has more samples. The investigation results provide a powerful tool for searching new 3CLpro enzyme inhibitors for SARS-CoV. © 2022, Publishing House of the Romanian Academy. All rights reserved.