Mechanistically driven identification of novel structural alerts for mitochondrial toxicity
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Abstract
Mitochondrial toxicity is a problem of growing concern in modern society, resulting in several serious adverse health effects including cardiac failure, hepatotoxicity, and neurodegenerative disorders. Adverse outcome pathways provide models for structuring a mechanistic understanding of toxicology, but most existing in silico models for prediction of mitochondrial toxicity do not account for the mechanism of action of potential toxicants. In a recent study by Hallinger et al., the Seahorse respirometric assay provides experimental data relating to the mechanism of action of several mitochondrial toxins. This makes possible, for the first time, the development of structural alerts for mitochondrial toxicity linked to mechanisms. By using a maximal common substructure searcher and Bayesian statistical analysis, we have discovered 11 alerts associated with different mechanisms of action. Eight of these are completely novel. By incorporating the mechanisms into the structural alert, more information can be gained about the molecular initiating events involved and build toward more complete adverse outcome pathways for mitochondrial toxicity.