thumbnail

An expert system for prediction of chemical toxicity

Drug Information Journal

By:
, , , and
DOI: 10.1177/009286159202600405

Links

Abstract

The National Fisheries Research Center- Great Lakes has developed an interactive computer program that uses the structure of an organic molecule to predict its acute toxicity to four aquatic species. The expert system software, written in the muLISP language, identifies the skeletal structures and substituent groups of an organic molecule from a user-supplied standard chemical notation known as a SMILES string, and then generates values for four solvatochromic parameters. Multiple regression equations relate these parameters to the toxicities (expressed as log10LC50s and log10EC50s, along with 95% confidence intervals) for four species. The system is demonstrated by prediction of toxicity for anilide-type pesticides to the fathead minnow (Pimephales promelas). This software is designed for use on an IBM-compatible personal computer by personnel with minimal toxicology background for rapid estimation of chemical toxicity. The system has numerous applications, with much potential for use in the pharmaceutical industry

Additional Publication Details

Publication type:
Article
Publication Subtype:
Journal Article
Title:
An expert system for prediction of chemical toxicity
Series title:
Drug Information Journal
DOI:
10.1177/009286159202600405
Volume
26
Issue:
4
Year Published:
1992
Language:
English
Publisher:
SAGE Publications
Publisher location:
Thousand Oaks, CA
Contributing office(s):
Great Lakes Science Center
Description:
7 p.
Larger Work Type:
Article
Larger Work Subtype:
Journal Article
Larger Work Title:
Drug Information Journal
First page:
487
Last page:
495
Number of Pages:
7