Desenvolvimento de modelos de predição de atividade inibitória sobre a DNA Girase de micobactérias baseados em estudos de modelagem molecular

Abstract

Tuberculosis (TB) is a disease with high morbidity and mortality. In 2019, approximately 10 million people developed the disease, of which about 1.4 million people died. There are effective antibiotics against the bacteria that cause tuberculosis, Mycobacterium tuberculosis (MTB), but the increased incidence of TB cases resistant to these drugs has required efforts to discover new drugs capable of fighting the disease. Topoisomerases are enzymes that maintain DNA topology during replication, transcription and recombination. DNA gyrase is the only type II topoisomerase present in MTB, which is why it is an interesting target to be explored for the design of new drugs against TB. DNA gyrase is composed of two subunits, GyrA and GyrB and this project aims to use groups of compounds from the literature with activity data recorded on the GyrB subunit of mycobacterial DNA gyrase, in order to develop a model for the prediction of activity of MTB DNA gyrase inhibitors to be applied in future virtual screening procedures. To achieve this goal, molecular modeling and multiple linear regression methods were combined to build prediction models of inhibitory activity (pIC50) for series of compounds present in the literature with inhibitory data on the inhibition of GyrB of M. smegmatis DNA gyrase. Good models were obtained, verified through internal validation with the LOO (Leave One Out) cross validation method. The results of the cross-validation were expressed by the correlation coefficient of the cross-validation (Q2) and the standard deviation of the cross-validation (SPRESS). The model's prediction statistics are expressed by the multiple correlation coefficient R2

EXT and the root mean square of the

prediction error (RMSEP).