Multivariate Characterization of Amino Acids by Reversed Phase High Pressure Liquid Chromatography

The twenty coded and four non coded amino acids were multivariately characterized by HPLC. The retention times for dansyl derivatives of the amino acids in nine eluents on a reversed phase C 18 column were used. With the nine HPLC variables and five additional variables (molecular weight, pK‐COOH, pK‐NH 2 , pI and substituent van der Waals volume) comprizing the matrix X, two partial least squares projections to latent structures (PLS) models were developed for eighteen of the coded amino acids, with the previously developed principal properties (PP.s) z 1 and z 2 as single y‐variables. The PLS models described 90% and 87% of variance in z 1 and z 2 respectively. Values for z 1 and z 2 for the four non coded amino acids were predicted from their X‐data and the PLS models.