In silico modeling of ecotoxicological hazards of chemicals using advanced chemometric tools
This is one out of several project ideas which were sent to New Indigo by scientists from India and Europe. If you would like to view the contact details of the provider of this idea, you have to log into your New Indigo account!
The experimental determination of environmental parameters of commercial chemicals is costly and time consuming. Since there is large number of chemicals currently in common use and new chemicals are registered at a very high rate, it is obvious that resources are insufficient to obtain experimentally even basic information on the environmental fate and effects of all these chemicals. Currently, there are ecotoxicological data available for less than 1% of industrial chemicals. Thus, it is necessary to develop quantitative models that accurately and readily predict environmental behaviour of large sets of chemicals.
A chemistry approach to predictive toxicology relies on quantitative structure-activity relationship (QSAR) modeling. The REACH regulation, recently implemented in the EU, recommends the use of valid QSARs for predicting the ecotoxicological properties of chemicals, in the interest of time- and cost-effectiveness and animal welfare. In the proposed project, in silico ecotoxicological models will be developed based on available toxicity data of chemicals (including pharmaceuticals) against various endpoints. It is very important that quantitative structure–toxicity relationships (QSTRs) obey the OECD principles. Our work emphasises external validation, partially through new and stringent statistical measures. We combine this attitude with two strengths: rigorous physicochemical parameters and innovative use of advanced machine learning.
Quantum topological molecular similarity (QTMS) indices will be used as electronic descriptors because they are compact and directly linked to modern quantum wave functions, which computers can now produce in large numbers. Kriging, a powerful up-and-coming method in QSAR, will benefit our in silico ecotoxicology, including daphnia, fish, Tetrahymena and reproductive toxicity, skin and eye irritation, mutagenicity, biodegradation and bioconcentration, and the largely unexplored ecotoxicity of pharmaceuticals.
- Earth and related environmental sciences