Course on in silico methods in the safety assessment of chemicals
The course covers various topics related to using in silico (computer-based) methods to assess the safety of chemicals. It provides training on machine learning methods, databases, molecular descriptors, similarity analysis, read-across approach, and QSAR/QSPR modeling. These methods are widely used in the assessment of chemical properties, toxicity, and potential risks associated with chemical substances.
The course is in English (or Polish – if the whole group of participants will be Polish-speaking).
- For (eco)toxicologists and chemists, risk assessors, regulatory affairs consultants, students, R&D specialists.
*Price includes: participation in classes, training materials, certificate of participation, coffee catering.
Programme
Day 1.
This workshop equips participants with knowledge of databases and effective search strategies to find relevant information. Additionally, participants will be introduced to two key unsupervised chemometric techniques: Hierarchical Cluster Analysis (HCA) and Principal Component Analysis (PCA). We will explore a methodology for building classifiers that predict crucial properties for risk assessment. These advanced techniques empower participants to analyze large datasets and present their findings visually.
Morning session:
- Introduction to machine learning methods in safety assessment of chemicals
- Introduction to Digital Chemistry and Advanced Nanoinformatics
- Databases and data FAIRnes
- Overview of database structure, search methods, types of information
- Examples of databases (PubChem, ChemSpider)
Afternoon session:
- Description of a chemical structure and molecular descriptors
- Introduction of descriptor’s group
- Practical exercises – calculation of Lipinski descriptors and 2D/3D descriptors using RDiK package.
- Similarity analysis: Introduction to HCA and PCA with practical exercises in KNIME
- Predict risk-relevant properties using a classifier’s.
Day 2.
Participant will be gradually introduced to building QSAR/QSPR models. They will learn about the essential data required for this process and the criteria that must be met to utilize the models for regulatory purposes. Apart from the modeling methodology, participants will also learn the practical use of available tools, allowing them to predict the physicochemical and toxicological properties of chemical compounds. These skills will empower them to make more informed decisions regarding the safety assessment of chemical substances.
Morning session:
- Grouping and read-across approach.
- Assumptions of the read-across method
- Criteria of substance similarity
- Read-Across Assessment Framework
- QSAR/QSPR modeling.
- Assumptions of the QSAR/QSPR modeling
- Phases of modeling (calibration, validation, applicability domain, application of the model)
- QMRF reports
Afternoon session:
- Application of in silico methods in risk assessment of chemicals.
- Practical exercise (1): Demonstration of the steps involved in predicting the physico-chemical properties / biological activity of a chemical based on QSPR/QSAR models
- Introduction to open-source tools to predict the properties of compounds.
- Practical exercise (2): Use of open-source tools for predicting the physico-chemical properties / biological activity of chemicals (VEGA / T.E.S.T.)
- Discussion of obtained results (comparing results depending on the tool used, searching for experimental data).
- Integrating in silico models and read-across methods for predicting toxicity of chemicals.
- All attendees will receive a course completion certificate.
- Better together - get a 10% discount on the open course for each additional employee from your company.
