Multiple regression analysis with optimal molecular descriptors

We consider construction of optimal molecular descriptors to be used for mu ltiple regression analysis of several properties of alcohols. The descripto rs are obtained by considering shorter paths with variable weight x for car bon-oxygen bond in alcohol. In particular we consider as molecular descript ors paths of length 1, 2 and 3. The multiple regression analysis of the fol lowing molecular properties was examined: - log S (S = solubility), CSA (ca vity surface area), log P (P = octanol/water partition), and log gamma (gam ma = infinite solution activity coefficient). By minimizing the standard er ror of the regression for each property we found optimal variable weight.