- Publisher:Lhasa Limited
- Publication Date:Sep 2016
- Publication Type:Poster
- Scientific Area:
- Industry Type:
Local electrophilicity as a descriptor for toxicity-relevant reactivity: A fragment-based approach
Presented by Dr. David Ponting at the 2016 EuroQSAR
A method for the rapid generation of local electrophilicity values for a set of compounds as described by (Wondrousch, et al., 2010) is presented. The dataset used for comparison is a list of 854 electrophilic compounds with associated in chemico reactivity data towards biologically-relevant nucleophiles from the review of (Schwobel, et al., 2011), curated to remove mixtures, ambiguous structures and counterions. These local electrophilicity values are then used to estimate the reactivity with the various nucleophiles investigated.
The Derek Nexus alerts of Lhasa Limited (Leeds, UK), which identify by means of pattern-matching the atoms of concern for various toxicological endpoints (including, but not limited to, skin and respiratory sensitisation, mutagenicity, carcinogenicity and hepatotoxicity) are used to mark the atoms of concern and separate the electrophiles into broad reactivity classes (e.g. Michael acceptors).
A Knime workflow is used for the rapid extraction of data from the output of GAMESS calculations (B3LYP/6-31+G**), generating an XML file with every single atom tagged with its appropriate charge- and energy- weighted electrophilicities, allowing the use of rapid, even combinatorial, regression analysis to determine the atoms whose properties are most significant, and unique parameters created for each of the broad reactivity classes.
Directed fragmentation methods are then used to fragment the molecules within each class, growing from the atoms indicated by the Derek Nexus alert. This creates a library of fragments and the relative effects on the electrophilicity of each one are then calculated. The set of fragments is then used to predict the local electrophilicity of novel compounds, not listed within the dataset of (Schwobel, et al., 2011) as well as to predict the reactivity of compounds within the dataset towards nucleophiles they have not been tested against.
Schwobel, J. A. H. et al., 2011. Measurement and Estimation of Electrophilic Reactivity for Predictive Toxicology. Chemical Reviews, Volume 111, pp. 2562-2596.
Wondrousch, D. et al., 2010. Local Electrophilicity Predicts the Toxicity-Relevant Reactivity of Michael Acceptors. Journal of PHysical Chemistry Letters, Volume 1, pp. 1605-1610.