Description:
Due to the importance of amino acids (AAs) as the basic bricks of proteins and their application in the drug and food industries, there is great interest in their separation and identification using simple and inexpensive approaches. Application of predictive models for the determination of the behavior of AAs can reduce trial-and-error experiments. Herein, the retardation factor (RF) of 21 protein AAs were studied using the quantitative structureretardation factor (QSRR) model. The RF values of the AAs in ethanol?sodium azide solution as the mobile phase of reversed phase thin layer chromatography (RP-TLC) were correlated with the structural properties of the AAs. The suggested QSRR indicated excellent fitting and prediction ability (R2train = 0.95 and R2test = 0.94). Furthermore, other statistical tests, such as y-scrambling, cross validation and the Williams plot confirmed the stability, absence of chance and the suitable applicability domain, respectively. It was shown that the sum of geometrical distances between oxygen and nitrogen atoms in an AA molecule is an important factor for the RF values of the AAs in the ethanol? sodium azide.