Development of new descripts for quantitative structure : activity relationship using fragrance compounds / by Assia Kovatcheva

• VerfasserIn: Kovatcheva, Assia
• Place / Publishing House: 2004
• Year of Publication: 2004
• Language: English
• Physical Description: VIII, 155 Bl.; Ill., graph. Darst.
• ac_no: AC04095534
• Hierarchical level: Monograph
• Statement of Responsibility: by Assia Kovatcheva

eng: Chirality is a topic of extended research due to the importance in stereoselective syntheses and stereospecific analysis of chiral molecules and particularly of biological reactions and interactions. Up to now, the enantioselectivity in several physiological and pharmacological processes has been established. However, the phenomenon of chirality in living organisms is still far from clear. The development of novel three-dimensional (3D) quantitative structure-activity relationship (QSAR) methodologies addressing the problem of stereochemical complexity leads to the generation of models which contribute to better understanding of the molecular mode of action as well to the successful target search in virtual databases of biologically active compounds. To this end, fragrance compounds were chosen as (i) they have complex stereochemistry; (ii) their mechanism of action is not yet understood; and (iii) it is an easily accessible testing system to evaluate new methods. Fragrance chemicals are a group of compounds, which allow the recognition of the biological effect in a fast and easy way. A novel 3D-QSAR approach that addresses the stereospecificity of a-campholenyl derivatives with sandalwood odor has been developed. This method does not require spatial alignment of molecules. The compounds of the data set are represented as derivatives of several common structural templates with several substituents, which are numbered according to their relative spatial positions in the molecule. Both wholistic and substituent descriptors were used as independent variables. A highly predictive QSAR models has been obtained and validated on an external test set of compounds (i. e. not used in the development of a QSAR model). The method used to assign substituents, can be applied to other data sets with complex stereochemistry. The model contributes also to the better understanding of structural, electronic and lipophilic properties of sandalwood fragrance compounds. To establish the efficiency of different representations of molecular chirality in QSAR studies, a combinatorial QSAR approach has been applied to a dataset of ambergris fragrance compounds with complex stereochemistry. The Combi-QSAR approach explores all possible combinations of different descriptor sets and correlation methods to obtain statistically significant models with high internal (for the training set) and external (for the test set) accuracy. Descriptor sets included MOE, CoMFA, CoMMA, Dragon, VolSurf, MolconnZ and 2D-chirality descriptors. Correlation methods included k-nearest neighbors (kNN) classification, Support Vector Machines (SVM), decision trees and binary QSAR. kNN classification in combination with CoMFA descriptors was found to be the best QSAR approach overall since predictive models were obtained for all divisions of the ambergris dataset into the training and test sets.