Franklin BAUER obtained his Master’s and Engineer’s degree in organic, bioorganic and therapeutic chemistry from National Graduate School of Chemistry of Mulhouse (ENSCMu) in 2014. During his studies, he gained a strong knowledge of organic chemistry and also learned molecular modelling, software programming and chemical analysis methods. Between 2014 and 2017, he did a PhD fellowship in collaboration with KREATiS and University of Haute-Alsace, focussing on the development of MechoA model to predict mechanisms of toxic action. Since then, Franklin is working at KREATiS on various QSAR models and software development.
OpenTox Summer School 2026
MechoA+ profiler, a new Add-in to the OECD QSAR Toolbox to predict mechanisms of action
The rise in interest for new approach methodologies (NAMs) has led to the development of a wide range of in silico approaches. Among them, KREATiS, Liverpool John Moores University (LJMU) and Unilever SEAC recently developed the MechoA+ scheme, by integrating and further refining the MechoA and Sapounidou-Firman schemes into one unique scheme. This tool is designed to help comprehend the mechanisms of toxic action related to chemical substances. It detects if the substance has a structural pattern known to be related to particular MIE(s) and for which species. Thus, it gives useful information for forecasting, controlling, and reducing the effects of chemicals on both human health and the environment.
The new in silico profiler is organised by molecular initiating events (MIEs) which are the first key events in Adverse Outcome Pathways (AOPs). Each mechanism is supported by (eco)toxicological data/ knowledge from the literature. The domain of each alert of the profiler has been defined in terms of structural features responsible for the MIE and species affected. In total there are more than 150 structural alerts for various MIEs. MechoA+ alerts have been coded as a new profiler for the OECD QSAR Toolbox. It has also been coded into iSafeRat® Desktop, providing the basic requirements for allocating a chemical to mechanistically-related QSARs. In order to test the relevance of the new scheme coded in MechoA+, the SMILES strings of over 70,000 structures were screened by MechoA+ and a comparison with results from the previous MechoA and Sapounidou-Firman schemes was undertaken.
The new MechoA+ scheme shows excellent coverage, with very few compounds unassigned, and a good accuracy compared to the mechanisms inferred from literature analysis. The evaluation demonstrates the comprehensive utility of the MechoA+ scheme, while also identifying some specific areas of chemistry where mechanistic assignment is currently lacking. In particular a need was identified to create further alerts for endocrine disruption.
In most cases, the results from the new scheme outperform the previous schemes in terms of statistics, wider applicability domain, and additional knowledge (e.g., species applicability, specific target hit, further key events, etc.), enhancing its usefulness for hazard assessment and understanding as well as confidence in the prediction.