A partial least squares and principal component regression study of quinone compounds with trypanocidal activity

Abstract

A quantitative structure–activity relationship (QSAR) study of 19 quinone compounds with trypanocidal activity was performed by Partial Least Squares (PLS) and Principal Component Regression (PCR) methods with the use of leave-one-out crossvalidation procedure to build the regression models. The trypanocidal activity of the compounds is related to their first cathodic potential (Ep_c1). The regression PLS and PCR models built in this study were also used to predict the Ep_c1 of six new quinone compounds. The PLS model was built with three principal components that described 96.50% of the total variance and present Q ²= 0.83 and R ² = 0.90. The results obtained with the PCR model were similar to those obtained with the PLS model. The PCR model was also built with three principal components that described 96.67% of the total variance with Q ² = 0.83 and R ² = 0.90. The most important descriptors for our PLS and PCR models were HOMO-1 (energy of the molecular orbital below HOMO), Q4 (atomic charge at position 4), MAXDN (maximal electrotopological negative difference), and HYF (hydrophilicity index).