Prediction of gas-chromatographic retention indices using topological descriptors

Theoretical prediction of gas-chromatographic retention indices could be us ed as an additional method for the identification of organic substances dur ing gas-chromatographic separation. Our previously developed model, based o n artificial neural networks, has been extended with the additional topolog ical structural descriptors to improve prediction capabilities. The topolog ical indices were selected for the representation of chemical structures be cause of their simplicity; therefore they could also be used for solving id entification problems by chromatographers who are not experts in structural representation. An extensive data set of 381 simple organic compounds with known retention indices taken from the literature served as a training and test set. Sixteen informational and topological structural descriptors wer e selected for the description of molecular structure. The same data set wa s used for the prediction of gas-chromatographic retention indices using a multiple linear regression model and back-propagation of error and counterp ropagation artificial neural network. The average root mean squared error v alues of a 10-fold cross-validation procedure were 22.5, 19.2, and 36.1, re spectively.