thumbnail

An expert system for prediction of aquatic toxicity of contaminants

By:
, , , and
Edited by:
Judith M. Hushon
DOI: 10.1021/bk-1990-0431.ch007

Links

Abstract

The National Fisheries Research Center-Great Lakes has developed an interactive computer program in muLISP that runs on an IBM-compatible microcomputer and uses a linear solvation energy relationship (LSER) to predict acute toxicity to four representative aquatic species from the detailed structure of an organic molecule. Using the SMILES formalism for a chemical structure, the expert system identifies all structural components and uses a knowledge base of rules based on an LSER to generate four structure-related parameter values. A separate module then relates these values to toxicity. The system is designed for rapid screening of potential chemical hazards before laboratory or field investigations are conducted and can be operated by users with little toxicological background. This is the first expert system based on LSER, relying on the first comprehensive compilation of rules and values for the estimation of LSER parameters.

Additional Publication Details

Publication type:
Book chapter
Publication Subtype:
Book Chapter
Title:
An expert system for prediction of aquatic toxicity of contaminants
DOI:
10.1021/bk-1990-0431.ch007
Volume
431
Year Published:
1990
Language:
English
Publisher:
American Chemical Society, ACS Symposium Series
Publisher location:
Washington, D.C.
Contributing office(s):
Great Lakes Science Center
Larger Work Type:
Report
Larger Work Subtype:
Other Report
Larger Work Title:
Expert systems for environmental applications