Suppr超能文献

Development of structure-taste relationships for thiazolyl-, benzothiazolyl-, and thiadiazolylsulfamates.

作者信息

Spillane William J, Coyle Catherine M, Feeney Brendan G, Thompson Emer F

机构信息

Department of Chemistry, National University of Ireland, Galway, Ireland.

出版信息

J Agric Food Chem. 2009 Jun 24;57(12):5486-93. doi: 10.1021/jf9002472.

Abstract

A total of 28 new five-membered aromatic ring thiazolyl-, benzothiazolyl-, and thiadiazolylsulfamates, as their sodium salts, have been synthesized and combined with 30 known similar heterocyclic sulfamates to create a database for the study of structure-activity (taste) relationships (SARs) in this heterocyclic subgroup, which is known to contain a somewhat disproportionate number of sweet compounds compared to other groups of tastants. A series of nine parameters (descriptors) to describe the properties of the sulfamate anions were calculated in Spartan Pro and HyperChem programs. These are the highest occupied molecular orbital (HOMO), lowest unoccupied molecular orbital (LUMO), length of the molecule, dipole moment, area, volume, E(solv), sigma (from the literature), and log P. The taste data for all 58 compounds were categorized into three classes, namely, sweet (S), nonsweet (N), and nonsweet/sweet (N/S). Discriminant analysis only classified 44 of the 58 compounds correctly. Classification and regression tree analysis (CART) using the S_ Plus program proved highly effective, in that the derived tree correctly classified 46 compounds from a training set of 48 and, from a computer randomly selected test set of 10 compounds, 7 had their taste correctly predicted. A second tree was grown using the additional taste category N/S, and this tree also performed extremely well, with 8 of the 10 compounds in the test set correctly classified. These trees should be very reliable for predicting the tastes of other heterocyclic sulfamates, which belong to the subset used here.

摘要

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验