A QSAR-BASED BIODEGRADABILITY MODEL - A QSBR

A microbial biodegradability predictive model has been developed using groups and molecular indices as molecular descriptors. The model cont ains 12 variables. The model was calibrated and later validated with a data set developed from biodegradation studies of acclimated activate d sludge metabolism. It also successfully predicted the biodegradation rate of several other substances with one exception within the limits established by a study of a second database. This fatter study was de signed to evaluate the expected scatter in acclimated biodegradation r ate data. The nitro group, amines, halogens and heteroatoms severely d ecrease the rate as do structural complexity and electronic alteration s created by most substituents. The carboxyl group markedly enhances t he rate but the hydroxyl group has little effect. It appears that the hydroxyl group is significant in acclimation but not in degradation. A discussion of statistical methods and software is included in this wo rk. Copyright (C) 1996 Elsevier Science Ltd