Retention data from affinity high-performance liquid chromatography in view of chemometrics

A combination of affinity chromatography and chemometrics is demonstrated to provide information on drug analytes and on biomacromolecules forming stationary phases, which is of relevance to molecular pharmacology and to rational drug design. The approach can also be applied to elucidate the molecular mechanism of enantioseparation on natural biopolymer stationary phases. Affinity high-performance liquid chromatographic data, which were determined on silica-based human serum albumin, alpha(1)-acid glycoprotein, keratin, collagen, melanin and amylose tris(3,5-dimethylphenylcarbamate) stationary phases, are discussed. Quantitative structure-retention relationships (QSRRs) derived for test series of drug analytes are interpreted in terms of structural requirements of specific binding sites on biomacromolecules. A means to quantify the differences in drug-biomacromolecule binding among the members of analyte families is demonstrated based on hydrophobicity and structural descriptors from molecular modeling. Chemometric processing of appropriately designed sets of affinity chromatographic data may increase the speed and efficiency of a search for new drugs, providing at the same time a chance to reduce the number of in vivo screenings. It can also be of help in rational selection of chiral columns for specific analytical separations. (C) 1998 Elsevier Science BN. All rights reserved.