Improvement in the in silico assessment of toxicological endpoints using Derek

23 August, 2018

The use of computer based (in silico) models for the evaluation of environmental fate, eco- and toxicological properties of chemicals is nowadays a well-established approach and also recommended within existing regulatory frameworks (e.g. biocides, industrial chemicals, agrochemicals and cosmetics…). These models are used to predict substance properties without testing.

To further strengthen our possibilities in the area of in silico methods within knoell, we have implemented Derek for Windows. This will improve the evaluation of many properties relevant for human health hazard and risk assessment. Derek for Windows is a well-known and widely used knowledge-based expert system developed by Lhasa Ltd. ( It includes rule-based models for the qualitative evaluation of toxicity endpoints. Derek models are based on well-documented structural alerts related to the properties of interest. The toxicity can be predicted for selected organisms (e.g. mammals, human, rat, dog, bacteria, etc.). The most common use of Derek is screening for potential mutagenicity or skin sensitization. However, several other endpoints are available as well, including carcinogenicity, irritation, organ toxicity, reproductive toxicity, respiratory sensitization and others.

The use of multiple models and the interpretation and evaluation of their outcomes by scientific experts in the relevant regulatory context is the heart of our computer-based approach for the assessment of toxicological properties. Derek for Windows supplements and strengthens the current QSAR evaluation strategy, and hence further increases the reliability of our model-based chemical safety assessments.

With the addition of Derek, we at knoell have implemented a valuable supplement to our models already in use which also widens our portfolio of QSAR services. Because of the above mentioned points, our reports further increase their already very high quality and reliability.

Business Units: