|HOME SCHEDULE AUTHOR INDEX SUBJECT INDEX|
(PW025) A comparison of QSAR models of biodegradation with OECD guidelines 301 data: relevant issues for the validation of these models.
Reuschenbach, P1, Silvani, M1, Dimitrov, S2, Mekenyan, O2, 1 BASF AG, Department of Product Safety, Ludwigshafen, Germany2 University Prof. Assen Zlatarov, Bourgas, Bulgaria
ABSTRACT- There is a need to develop QSAR based approaches that are applicable for broad chemical classes to address the escalating demand for information on their potential effects on the environment in the context of, for example the European Commission White Paper and High Production Volume programs (HPV). Since the knowledge of the fate of a substance is a central point for an environmental assessment, two QSAR models for biodegradation, BIOWIN and CATABOL, were analyzed with respect to their reliability, applicability and uncertainty. The BIOWIN system contains six models and predicts the probability of ready biodegradability and the range of half lives. CATABOL calculates the extent of degradation with confidence limits in ready tests. Moreover, it simulates microbial degradation pathways in order to identify potential metabolites. To assess the model performance, biodegradation data from BASF were randomly collected and evaluated for ca. 480 substances with widely varying chemical structures. All data were determined according to the OECD test guidelines for ready biodegradation (OECD 301 A-F). Substances were judged either readily biodegradable or not readily biodegradable. The BIOWIN and CATABOL predictions were compared with experimentally derived data to analyze: (1) the overall accuracy of the models to predict biodegradation, (2) the effect of different published pass level criteria for evaluating readily tests and (3) the impact of the availability biodegradation data from different OECD 301 tests on the correlation to the QSAR data. Based on the results of this study it was concluded that the QSAR models should only be used by experts who can objectively assess their performance on substances in question.
Key words: Biowin, QSAR, Catabol
Internet Services provided by|
Allen Press, Inc. | 810 E. 10th St. | Lawrence, Kansas 66044 USA
e-mail firstname.lastname@example.org | Web www.allenpress.com
All content is Copyright © 2004 SETAC