RECETO PhD course: Use of QSAR models - practical use of the CAESAR models for legislative purposes (e.g. REACH legislation) - 07.07.2011
Use of QSAR models:
practical use of the CAESAR models for legislative purposes (e.g. REACH legislation)
Type of course: introductory PhD course. Short course (3 hours – with the possibility to include 2 hours of voluntary exercises).
Target participants: Academic: senior, Post Doc, PhD but also for interested MSc students, regulators and industries.
Participants have to bring their own laptop to the course.
DEADLINE for REGISTRATION: 20 July 2011, MAXIMUM 15 participants will be accepted, please fill out the registration form and send it with the required attachment to .
Date of the course: 23 August 2011.
Fee: there is no course fee.
Credits: 1 ECTS is available for students who perform an written assignment after the course (~25 hours work load). The assignment will consist of a small report including analysis using the methodology from the course. The analysis may be performed on own data or examples provided by the teacher. To be approved the assignment must show a good level of comprehension of the explained topics.
Where: Faculty of Life Sciences at Copenhagen University. Room A1-02.01 on Nordre Sti, Frederiksberg Campus (see attached map of Frederiksberg Campus).
Course responsible: Dr Elena Boriani (Mario Negri Research Institute, Italy) / NERI, Denmark.
Description: Within the CAESAR EU project (http://www.caesar-project.eu) specific in-silico models have been developed to predict the biological activity of chemicals for legislative purposes, and further improved within other projects (http://www.orchestra-qsar.eu, http://www.antares-life.eu/, ..). This course provides a specific practical overview of these QSAR models, but also regarding QSAR models and predictive software in general. Particular attention will be given to the issue of the applicability domain. The course will be mainly practical, addressing the typical errors which can occur with these methods, and how to avoid them, in order to have safer results. This course is strongly recommended to understand the meaning of QSAR model results and to critically and correctly interpret them in the context of chemical safety for risk assessment purposes. The new platform VEGA, developed in collaboration with the US EPA and other regulatory bodies, will also be introduced. It makes available the CAESAR models and others, with automatic exploration and visualization of the results, providing a measurement of the model reliability.
Objectives: The course will provide examples and guidance on the practical use of particular QSAR, for assessing chemical properties and toxicity data for compounds. CAESAR models will be tested and results will be critically discussed. The course is not aiming to teach all QSAR features but recommended practices to deal with QSAR methods in a critical way. Participants will learn how to use QSAR models, what can be obtained and what it cannot. After the course students will be able to perform their own calculations using the user friendly, freely available CAESAR models. They will also be able to interpret the obtained results.
Course outline: Different possible uses of QSAR for risk assessment and legislations (e.g. registration, classification and labeling, prioritization). Introduction to the approach given by legislations (e.g. REACH) on the QSAR methods. Introduction to the CAESAR QSAR methods and new free platform VEGA for QSAR modeling. Presentation of the five endpoints in CAESAR: Carcinogenicity/BCF/mutagenicity/skin sensitization/reproductive toxicity. Results from the validation sets for these models will be shown, and how to use them. The tool for evaluation of the applicability domain available within CAESAR will be introduced. The evaluation will be addressed according to the OECD principles: accuracy, sensitivity, specificity, false positives and false negatives. Pitfalls of the QSAR methods will be shown, discussing when not to use the models, and how to take advantage of different models. Participants will use the models themselves, through a series of different exercises. Participants can actively participate also using own data as case studies. The advantages and disadvantages of the different QSAR methodologies will be discussed. Comparison within experimental values /other software results. Understanding the meaning behind a QSAR approach.
PROGRAM
9.00 - 9.20: Introduction to alternative methods and different possible uses QSAR of e.g. for REACH (registration, classification and labeling, prioritization)]
9.20 - 9.45: Brief introduction on the CAESAR QSAR methods.
The five endpoints in CAESAR will be presented:
• BCF
• Skin Sensitisation
• Mutagenicity
• Carcinogenicity
• Developmental Toxicity
Results for the validation sets using these models, and how to apply the models.
The tool for the evaluation of the applicability domain will be introduced. Evaluation will be addressed according to the OECD principles, accuracy, sensitivity, specificity, false positives and false negatives. Pitfalls of the QSAR methods will be shown, discussing when not to use the models, and how to take advantage of different models. Participants will use the models themselves, though a series of different exercises.
9.45 - 10.15: Coffee break
10.15-11.15: The new VEGA platform. How to analyse the information used by the model, for each individual prediction. The advantages and disadvantages of the different QSAR methodologies will be discussed.
• Comparison within experimental values /other software results
• Understanding the meaning behind a QSAR approach
• Advantages and disadvantages of the methods will be summarized.
11.15-11.45: Participants will use the models themselves, though a series of different exercises.
11.45-12.00: Review, questions and course evaluation.
12.00-13.00: Lunch (sandwiches will be provided).
13.00-15.00: The instructor will be available for further exercises, questions.
Course materials: Re-prints of relevant published articles, handouts of lecture slides and other course materials including files of the calculated examples and useful links.
Ditte Carlsen, - last update:12 July 2011