Prediction of gas chromatographic relative retention times of stimulants and narcotics

The ADAPT software system was used to create models for the prediction of gas chromatographic relative retention times (RRTs) of stimulants and narcotics that are analyzed in doping control of athletes. The two main methods that were followed for building the models were the quantitative structure-retention relationship (QSRR) and multiple linear regression analysis. The main proposed model for the entire data set had a multiple correlation coefficient R = 0.991 and standard error s = 0.046 or approximately 4.5%. Because of the relatively high standard error of the main model, a second model was built on a subset of compounds with R = 0.982 and s = 0.027 or approximately 2.5%.