Multivariate evaluation of the correlation between retention data and molecular descriptors of antiepileptic hydantoin analogs

• Published in: Journal of chemometrics 2012-03, Vol.26 (3-4), p.95-107
• Main Authors: Djakovic-Sekulic, Tatjana, Smolinski, Adam, Trisovic, Nemanja, Uscumlic, Gordana
• Format: Article
• Language: English
• Subjects: antiepileptic drugs; Chemical compounds; Chromatography; partial least squares; Prescription drugs; Principal components analysis; quantitative structure-retention relationships; Regression analysis; robust partial least squares; stepwise regression
• ISSN: 0886-9383; 1099-128X
• DOI: 10.1002/cem.1421

Molecular properties relevant to pharmacokinetics of two sets of newly synthesized hydantoin derivatives based on two well‐known drugs (Nirvanol and Phenytoin) were studied. Properties under consideration were either determined empirically by reversed‐phase high‐performance thin‐layer chromatography or calculated by the use of established theoretical medicinal chemistry/drug design software. These properties represent selected structural features of analytes, which affect their processes of absorption, distribution, metabolism, excretion, and toxicity. Principal component analysis was used to visualize the differences between six mobile phase modifiers used as well as similarities between 24 analyzed compounds. To find appropriate quantitative relationships between RM, W for the tested compounds and molecular descriptors, stepwise regression, partial least squares (PLS), and robust PLS were used. The best results were obtained with robust PLS. Copyright © 2012 John Wiley & Sons, Ltd. Molecular properties relevant to pharmacokinetics of 24 newly synthesized hydantoin derivatives based on two well‐known drugs, Nirvanol and Phenytoin, were studied. Properties under consideration represent selected structural features that affect their processes of absorption, distribution, metabolism, excretion, and toxicity. To find appropriate quantitative relationships between RM, W for the tested compounds and calculated molecular descriptors, principal component analysis, stepwise regression, partial least squares (PLS), and robust PLS were used.