Quantitative structure-activity relationship study of nonpeptide antagonists of CXCR2 using stepwise multiple linear regression analysis

The chemokine receptor CXCR2 plays an important role in recruiting granulocytes to sites of inflammation and has been proposed as an important therapeutic target. A linear quantitative structure-activity relationship model is presented for modeling and predicting biological activities of CXCR2 antagonists. The model was produced by using the multiple linear regression technique on a database that consists of 55 nonpeptide antagonists of CXCR2. Stepwise regression as a variable selection method was used to develop a regression equation based on 43 training compounds, and predictive ability was tested on 12 compounds reserved for that purpose. Appropriate models with low standard errors and high correlation coefficients were obtained. The mean effect of descriptors and standardized coefficients shows that the mean atomic van der Waals volume is the most important property affecting the biological activities of the molecules. The square regression coefficient of prediction set for the multiple linear regression method was 0.912.