We use cookies on our website. Some are necessary for the operation of the website. You can also allow cookies for statistical purposes. You can adjust the data protection settings or agree to all cookies directly.
Data Protection Declaration
Imprint
Further development of in silico models for prediction for mutagenicity of pesticides - (Q)SAR
Project
Project code: BfR-SiP-08-1322-
Contract period: 01.01.2018
- 31.12.2018
Purpose of research: Inventory & Assessment
During authorisation procedure of plant protection products, biocides and chemicals increasingly in-silico studies, especially (Q)SAR are submitted by applicants from idustry. These in-silico studies are based on computer-aided prediction models, which - by means of assumptions about structure activity relationships - predict toxicity properties of chemical substances. A major advantage of (Q)SAR is, besides cost and time savings, in particular the reduction and/or replacement of animal testing (RRR-principle). For regulatory requirements the establishment and application of (Q)SAR- methods is reflected in several guidance documents from ECHA, EFSA and OECD (1-6). For the BfR, as a regulatory working authority, the in silico programs for the prediction of toxicity are indispensable to follow and critically evaluate submitted studies by applicants. These in-silico tools assist to perform a risk assessment independent from industry and applicant data. In general, in-silico programs can be grouped into rule-based expert systems and quantatitative structure activity relationship systems (QSAR-systems). Relevant guidance documents (7,8) and scientific publications (9) recommend the combined use of both systems to minimize the number of false-positive or false-negative predictions. It became obvious during the last years that assessment of toxicity by industry is often conducted with the expert-system DEREK. Additionally, the use of the complementary QSAR-system SARAH fits into the recommended (Q)SAR approach. Both programs are distributed by the non-profit company Lhasa Limited. A number of five software-licences of the programs DEREK and SARAH for toxicity prediction are available at the BfR since 2017, thereof four in departement 6 of the BfR(Pesticides Safety). The toxicity prediction of DEREK covers numerous endpoints among others mutagenicity, whereas until now SARAH only makes predictions for the endpoint mutagenicity. In a pilot project it has been examined whether the prediction of mutagenicity from 30 selected pesticide-active substances and metabolites (among them 15 mutagenic and 15 non-mutagenic substances) match with the experimental test result. It could be shown that the mutagenic properties of only 6 out of the 15 selected mutagenic substances were predicted correctly. The reason for this might be based on the underlying database of toxicological structures (toxicophores) used for the prediction of mutagenicity. It is assumed that the number of pesticide data in the dataset is insufficient which in turn results in a limited number of structural alerts needed for flagging mutagenic substances. Therefore, a key goal of this project will be a systematic evaluation, further development and adaptation of DEREK and/or SARAH for their application in regulatory assessment of pestcides. Objective: The aim of this project is to examine whether the prediction for mutagenicity can be improved by implementing further mutagenicity data of active substances and metabolites from the plant protection and biocide assessment process into the database of DEREK and/or SARAH. Parameters therefore could be the concordence between experiment and prediction and therewith the reduction of false-positive or false-negative prediction-results.
Section overview
Subjects
- Crop Protection
- Toxicology
- Computer science
