Spillane W J, McGlinchey G, Muircheartaigh I O, Benson G A
J Pharm Sci. 1983 Aug;72(8):852-6. doi: 10.1002/jps.2600720804.
Eleven heterosulfamates have been synthesized, characterized, and evaluated for sweetness. Measurements of the molecular dimensions (x, y, z, and V) of these sulfamates and 22 others that had been reported previously and evaluated for sweetness have been made using Corey-Pauling-Koltun space-filling models. The first-order molecular connectivities (1Xv) of all the heterosulfamates have been calculated. The statistical technique of linear discriminant analysis was applied to the complete set of 33 compounds and to a reduced set of 27 compounds. The analysis was performed using the above five variables (x, y, z, V, and 1Xv) and various subsets thereof. For the complete set of compounds, seven variable subsets were identified which yielded correct classifications of 27 and 28 compounds. A similar analysis of the reduced set did not improve the misclassification rate.
已合成、表征并评估了11种杂氨基磺酸盐的甜度。使用Corey-Pauling-Koltun空间填充模型测量了这些氨基磺酸盐以及之前报道并评估过甜度的其他22种氨基磺酸盐的分子尺寸(x、y、z和V)。计算了所有杂氨基磺酸盐的一阶分子连接性(1Xv)。将线性判别分析统计技术应用于33种化合物的完整数据集和27种化合物的简化数据集。使用上述五个变量(x、y、z、V和1Xv)及其各种子集进行分析。对于完整的化合物集,确定了七个变量子集,它们对27种和28种化合物进行了正确分类。对简化集进行的类似分析并未提高错误分类率。