______________________________________________________________________________________________________________
QT-040
CHEMOMETRIC AND QSPR ANALYSIS OF A SET OF HYDROPATHY SCALES
Márcia M. C. Ferreira (PQ), Rudolf Kiralj (PQ),
Instituto de Química, Universidade Estadual de Campinas, Campinas 13083-970, SP, Brasil

Palavras Chave: hydropathy scales, semi-empirical PM3 method, chemometrics
A set of 88 hydropathy scales studied here was treated by Principal Component (PCA) and Hierarchical Cluster Analysis (HCA). Molecular geometries of 20 natural amino-acids and their residues were optimized at semi-empirical PM3 and ab initio level. Molecular descriptors for the amino-acids (hydrophobicity, dipole moment, molecular volume and surface area, number of non-s electrons, aromaticity vector) and their residues (hydrophobicity, dipole moment) were generated and related to the hydropathy scales via Principal Component Regression (PCR). The hydropathy scales are described well at least by four principal components (over 90% variance) in PCA. Both PCA and HCA show clustering of charged polar, uncharged polar, non-polar and essentially hydrophobic amino acids. The 88 PCR models (one for each hydropathy scale, with the same molecular descriptors) resulted in significantly different quality parameters (calc. r = 0.273 - 0.996).
______________________________________________________________________________________________________________
 

25a Reunião Anual da Sociedade Brasileira de Química - SBQ