Master's thesis: Are Mistra SafeChem Tools Regulatory Ready?

Doing your master's thesis with us at IVL gives you the opportunity to get a unique insight into our research and our work. It's a chance to learn a lot more in your field of study. We welcome thesis students every semester. One or two students are invited to perform their master's thesis fulfilling 30-60 credits (HP).

About the master thesis

The EU aims to develop a safe and sustainable chemical industry through its Chemicals Strategy for Sustainability and the Principles of Green Chemistry. Mistra SafeChem contributes to this goal by developing new production methods, tools for predicting chemical hazards and risks, and life cycle approaches. After a successful first phase (2019–2024), the programme continues in Phase II (2024–2028), with a focus on designing safer and more sustainable chemicals (Safe and Sustainable by Design, SSbD).

The programme combines green chemistry with modern methods, including computer-based models (in silico, such as AI and machine-learning-based QSAR models) and laboratory tests (in vitro), to predict hazards and exposure. These tools have already been tested in industrial settings. However, their relevance to regulatory frameworks such as REACH and CLP still needs to be evaluated systematically.

Importantly, these methods are intended to support, not replace, existing regulatory testing. They provide early insights that may guide further testing and could contribute to the development and future acceptance of New Approach Methodologies (NAMs).

Aims

This project focuses on evaluating the regulatory relevance of these methods through systematic data collection and comparison.

The student will compile and analyse available in vivo and higher-tier in vitro endocrine disruption (ED) data using regulatory databases, such as ECHA and the US FDA. This data will be compared with predictions from machine-learning models developed within the programme and from other open-sourced platforms (e.g. OECD QSAR Toolbox and VEGA).

The aim is to:

· assess how well model predictions align with regulatory data and decision criteria,

· identify gaps in relation to REACH/CLP information requirements,

· and evaluate how in silico approaches can support regulatory hazard assessment.

In addition, the project will examine key aspects of current regulatory frameworks, including how data are accepted and the role of New Approach Methodologies (NAMs). This will help improve understanding of how computational methods can complement established testing strategies.

Expected outcomes and methods

Methods

· Data collection from regulatory databases (e.g., ECHA, OECD, US FDA)

· Literature review of regulatory approaches to endocrine disruption assessment

· Comparative analysis of regulatory data and in silico model predictions

· Evaluation of agreement with REACH and CLP decision criteria

The project may also include basic data processing, visualisation, and statistical comparison of prediction performance.

Tools and Resources

The student will use publicly available regulatory databases, scientific literature, and in silico prediction tools developed within the Mistra SafeChem programme.

Relevant software may include:

· data analysis tools (e.g., Excel, KNIME, R, Python)

· cheminformatics or QSAR-based platforms

The choice of tools may depend on the student’s background and interests.

Expected Outputs

The thesis will provide a comprehensive evaluation of the regulatory applicability of in silico methods for predicting endocrine disruption. This includes comparing model outputs with available regulatory data.

The work is expected to:

· identify strengths and limitations of current approaches

· highlight data gaps

· contribute to the future use of NAMs in chemical hazard assessment

The results will be presented in a written master thesis and may also contribute to ongoing research within the Mistra SafeChem programme.

In addition, the project will give the student practical insight into EU chemicals legislation, particularly REACH and CLP, and hands-on experience with regulatory data. Through data collection, analysis, and comparison with model predictions, the student will build a strong foundation in regulatory hazard assessment and gain understanding of how in silico methods can be applied in a regulatory context.

Contact

For questions, please contact: Monika Witala monika.witala@ivl.se or Ziye Zheng ziye.zheng@ivl.se

Interested?

Submit your interest by clicking on "Registration of interest".