Graph Theoretical Atom-Type-Based Descriptors for Structural Characterization and Retention Prediction of Acyclic Alkanes
Fariba Safa, M. Jafari Ghadimi · Moscow University Chemistry Bulletin · 2021
Abstract The atom-type-based AI topological indices (TIs) along with the Xu index were utilized to generate the multiple linear regression (MLR) and artificial neural network (ANN) models for estimating the gas chromatographic retention index (RI) values of 155 normal and branched alkanes on the squalane stationary phase. The developed 5–8–1 ANN could more accurately predict the retention data than MLR. Based on the findings, relative importances of TIs on the retention indices decreased as AI(>C Xu > AI(–CH3) >AI(>CH–) >AI(–CH2–). The study indicated the significant role of the bulkiness as well as the position and degree of the molecular branching in determining RI values of the model compounds on the nonpolar stationary phase.