Posters
« Back
Novel QSAR Models for Predicting Toxicity of Chemicals to Aquatic Organisms and Identifying the Mode of Action
EP20652
Poster Title: Novel QSAR Models for Predicting Toxicity of Chemicals to Aquatic Organisms and Identifying the Mode of Action
Submitted on 20 Dec 2013
Author(s): Kiril Lanevskij, Liutauras Juska, Remigijus Didziapetris and Pranas Japertas
Affiliations: ACD Labs
Poster Views: 2,507
View poster »


Poster Information
Abstract: The utilized modelling approach was validated by applying the same principles to develop a predictive model for IGC50 (50% inhibitory growth concentration) to protozoan Tetrahymena pyriformis. This model was submitted as an entry for environmental toxicity prediction challenge hosted by CADASTER project. The model derived using known IGC50 values for 644 compounds was identified among the winners achieving RMSE under 0.8 log units for blind validation set of 120 chemicals. It is also demonstrated that the chemicals Mode of Action (MOA) can be determined using a simple set of structural fragments associated with certain MOA classes.
Summary: This study presents the application of recently introduced GALAS modeling methodology for estimating toxicity of chemicals to several aquatic species. Experimental data were expressed as median lethal concentration of test compound in water (LC50) and the data set contained 904 LC50 values for fishes (Pimephales promelas) and 589 LC50 values for crustaceans (Daphnia magna).
Report abuse »
Creative Commons