- Data modelling based on advanced AI algorithms.
- Prediction of structural alerts by quantitative structure-activity relationship models (QSAR).
- Prediction of toxicity/pharmacological properties and grouping of compounds based on read-across analysis.
- Prioritisation/grouping of chemicals and definition of libraries of chemical test items by virtual screening
- Definition of model intrinsic applicability domain by virtual screening.
![](https://research.tuni.fi/uploads/2021/05/0d8f595f-in-silico.png)