A new in silico platform that facilitates the evaluation of mutagenicity with regulatory validity launched

Mutagenicity (capacity to alter cellular genetic material) is one of the most important parameters for evaluating the possible toxicity of chemical products, particularly during their mandatory registration before accredited regulatory authorities. Up to now, mutagenicity has been detected in a standardized way using the Ames test, an in vitro test that uses genetically modified strains of the Salmonella typhimurium bacteria. Alternatively, computational methods (in silico) may be employed, and these are not only equally valid at regulatory level, but also much faster and cheaper.

The evaluation of mutagenicity in impurities of pharmaceutical products is governed by the International Council on Harmonization (ICH), and in particular by the ICH-M7 guide. According to this guide, computational toxicological evaluation is valid for regulatory purposes when two complementary methodologies are used to predict the outcome of a bacterial mutagenicity test: models based on "expert rules" and those of statistical nature (QSAR, "quantitative relationship between structure and activity”). The prediction of non-mutagenicity from these two complementary methodologies is sufficient to conclude that the impurity is not of mutagenic concern, and therefore no further testing is deemed necessary.

ProtoQSAR, a Spanish company specializing in chemoinformatics and molecular modelling, has recently launched ProtoICH, a computational tool that allows for mutagenicity predictions with regulatory validity in a very simple, fast and economical way.

ProtoICH integrates QSAR and rule-based models, which have been built using the largest reliable data set used to date for mutagenicity prediction, and allows users to predict the mutagenicity of their molecules of interest using the two methods as required by the ICH-M7 guide, by simply submitting the corresponding chemical structures (e.g. in the standard SMILES format, or even by drawing the molecule), or some standard identifier such as the CAS number.

ProtoICH presents outstanding advantages over currently available computational alternatives:

1. It has the highest predictive capacity described to date (with a proven degree of accuracy of around 85%, that is, a value similar to the best inter-laboratory reproducibility values of the Ames test).
2. While other computational tools are "black boxes" to the users -in the sense that the degree of certainty or reliability of their predictions may not be established, and only experienced personnel may work with these tools and interpret their results-, ProtoICH is extremely easy to use and provides the user with all the information about the models that comprise it.
3. Shows a wider applicability domain than other alternatives: while other tools have restrictions on the set of compounds they can reliably analyze, ProtoICH can be used for virtually any chemical compound.
4. Finally, ProtoICH allows one to automatically obtain the reports required at the regulatory level.

ProtoICH is the first module of ProtoQSAR's ProtoPRED in silico platform, which will allow for the prediction of a large set of properties of chemical compounds from their structures: ranging from physical-chemical properties to biomedical, therapeutic, pharmacological (such as ADME), toxicological properties and ecotoxicological, including those required by the REACH regulation.