Uncertainty Quantification for the Development and Evaluation of Statistical Methods in Next Generation Risk Assessment

• Applicant: Dr. Julia Duda
• Subject Area: Epidemiology and Medical Biometry/Statistics ; Toxicology, Laboratory Medicine
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 554317037

The aim of the planned project is to contribute to the development and use of new toxicity testing methods in order to ensure a sustainable risk assessment of new substances and thus the protection of public health. New substances are being developed ever faster. In order to protect public health, they are tested for toxicity according to complex regulatory guidelines before they are used by consumers or processed by workers in the industry. National and international regulatory ordinances require toxicity testing based on animal studies before a new substance can be approved. However, animal testing is time-consuming, expensive and can no longer cope with the growing number of newly developed substances. In addition, the quality of the transferability of toxicity results based on the physiology of animals to the physiology of is increasingly in question. In order to counteract the backlog of untested substances and to enable risk assessment of substances faster, more cost-effectively and geared towards human physiology, a paradigm shift towards next-generation risk assessment (NGRA) has started. NGRA utilizes the technological development of the last decades and combines high dimensional genetic data from different human cell-based test assays, existing knowledge on chemical properties and molecular mechanisms of action with appropriate statistical methods to quantify toxicity risks of unknown substances without the need for extensive animal testing. This results in two main challenges, both of which are addressed by this project. On the one hand, the statistical methods must be further developed so that the various data sources can be evaluated in optimal combination. Secondly, the inevitable complexity of these new methods poses a problem for the regulatory approval procedures. Accordingly, there are two goals for this project: to advance the statistical method development for NGRA methods and to facilitate the evaluation of complex NGRA methods. The first concrete goal of this project is to develop a new statistical method for flexible modeling of dose-response relationships - a key step in data modeling and evaluation for NGRA methods. The second aim is to develop a transparent evaluation approach based on the recently developed Bioactivity-Exposure-Ratio. This will quantify the uncertainty of new NGRA methods in a standardized way and thus make the approval process for NGRA methods more objective, simpler and safer. Both objectives make a significant contribution to ensuring the long-term safety of new products and food.

• DFG Programme: WBP Fellowship
• International Connection: USA
• Host: Nicole Kleinstreuer, Ph.D.