Quantitative structure-property relationship studies of the odor threshold of odor active compounds

Computer assisted statistical methods have been used to studythe odor thresholds of two sets of odor active molecules. Onedata set included 53 aliphatic alcohols. The second data setincluded 74 monoand di-substituted pyrazine derivatives. Eachquantitative structure-activity relationship study was executedin two steps. First, linear regression equations were developedwhich modeled the Ková;ts retention indices of both setsof compounds on two stationary phases, OV-101 and Carbowax-20M,with excellent diagnostic statistics (R > 0.99). Those resultswere critical in the second phase of each study which was thedevelopment of linear regression equations to calculate thelogarithm of the odor threshold of the compounds (R > 0.92).One significant feature of certain descriptors found to be importantin these regression models is that they highlight the size ofthe molecules.