The Institute of Food Safety, Animal Health and Environment “BIOR” continues the implementation of research application No. 1.1.1.9/LZP/1/24/124, “From Data to Decisions: Development of Smart Tools for Food Safety Risk Assessment and Prioritisation”.

The project focuses on the development of data-driven approaches and digital tools to support the assessment, comparison and prioritisation of chemical risks in food. In practical terms, the project addresses one of the central challenges in modern food safety risk assessment: how to combine occurrence data, food consumption data, toxicological reference values, scientific literature and predictive toxicology information in a structured and transparent way.

During the reporting period, work continued on the project’s data infrastructure, toxicological data integration, consumption data linkage, risk prioritisation methodology and international cooperation.

Data collection, harmonisation and toxicological reference values

During the reporting period, the first work package, “Data collection and harmonisation”, was completed. This included the compilation of a list of chemical contaminants relevant to the project, acquisition of contaminant occurrence data from Latvian food monitoring and other sources, and the standardisation, cleaning and preparation of these data for further integration into the project database.

This work forms the technical basis for the later stages of the project. For occurrence data to be used in exposure assessment and risk prioritisation, they must be comparable across food groups, chemical substances, analytical parameters, measurement units and data quality descriptors.

Work also continued under the second work package, “Compilation and application of health-based guidance values and predictive toxicology”. This part of the project focuses on collecting health-based guidance values, such as acceptable daily intakes, tolerable daily intakes and other toxicological reference values. These values are needed to interpret contaminant concentrations in food in relation to potential human health relevance.

At the same time, data gaps are being identified, including cases where full toxicological information is not available. For these cases, the project is assessing how predictive toxicology approaches and New Approach Methodologies can support future prioritisation and decision-making.

Integration of food consumption data and model development

The third work package, “Integration of consumption data and model development”, was initiated during the reporting period. The work focuses on how contaminant occurrence data can be linked with food consumption data and toxicological reference values.

This step is important because the presence of a contaminant in food does not by itself define the level of risk. Risk prioritisation also depends on how frequently and in what quantities the food is consumed, which population groups are exposed, and how the observed concentrations compare with toxicological reference points.

Full exposure modelling has not yet been completed, but the methodological basis for using consumption data in later risk prioritisation models has been prepared. During this work, an additional need was identified: a technical component for extracting and structuring contaminant occurrence data from scientific literature. This component has been conceptualised as Occurrence Router. Its purpose will be to support the identification, extraction and structuring of occurrence data reported in scientific publications, where relevant information is often distributed across tables, supplementary files and heterogeneous reporting formats.

Participation in the 3rd BeSafeBeeHoney Conference in Tirana

From 6 to 8 May 2026, the researcher participated in the 3rd BeSafeBeeHoney Conference, “Building a Safe European Value Chain in Beekeeping products”, held in Tirana, Albania. The conference was organised within COST Action CA22105 BeSafeBeeHoney and brought together researchers and stakeholders working on bee product safety, quality, authenticity, contaminants, biomonitoring and One Health-related aspects of apiculture.

Within the conference, the researcher participated in scientific sessions and discussions related to chemical contaminants in beekeeping products and presented work on glyphosate occurrence in beekeeping matrices, including honey, pollen and bee bread. These activities are closely connected to the postdoctoral project because bee products represent a relevant case study for occurrence data collection, literature-based data extraction, matrix harmonisation and risk prioritisation.

The conference also supported ongoing cooperation within Working Group 2 of COST Action CA22105, including work on the systematic review of persistent and emerging environmental contaminants in edible bee products. This work contributes to the broader methodological direction of the postdoctoral project by addressing how published occurrence data can be collected, standardised and interpreted for risk assessment purposes.

The BeSafeBeeHoney network is thematically relevant to the project because it connects expertise from chemistry, biology, ecology, veterinary science, beekeeping, agriculture, nutrition, economics and policy in relation to bee products and their value chain.

Scientific mobility at the University of Maribor

A major activity during the reporting period was the scientific mobility at the University of Maribor, Slovenia, carried out from 11 May to 7 June 2026 in cooperation with Prof. Tomaž Langerholc.

The mobility focused on the development of risk classification and prioritisation approaches for food safety. A central part of the work was the detailed analysis of the previously developed Risk Thermometer approach, originally developed with Swedish colleagues for risk prioritisation.

The Risk Thermometer approach is relevant to the postdoctoral project because it provides a structured way to compare and prioritise chemical risks. Instead of treating each contaminant or food group in isolation, the approach supports the development of a comparative framework where multiple substances, food categories and evidence sources can be evaluated together.

During the mobility, the technical version of the tool was rebuilt. Data input and output were optimised, and new visualisation and filtering functions were added. The updated version was tested using previously applied datasets, producing the same results as the original tool. This confirmed the functional consistency of the rebuilt version and created a stronger basis for further methodological development.

The mobility also supported the development of a joint review article concept on risk classification and prioritisation methods in food safety. In addition, the need for the Occurrence Router component was further clarified, especially in relation to the automated extraction of occurrence data from scientific literature.

As a next step, closer cooperation with Swedish colleagues was identified as important for the further development, adaptation and application of the Risk Thermometer methodology within the postdoctoral project.

Participation in the CHANGE workshop in Oslo

From 23 to 25 June 2026, the researcher participated in the international CHANGE workshop “Designing Interventions for Effective Use of NAMs in Regulatory Toxicology”, held in Oslo, Norway.

The CHANGE initiative focuses on system-level interventions to support the effective use of New Approach Methodologies in regulatory toxicology. Participation in the workshop was relevant to the project not only from the perspective of predictive toxicology, but also from the perspective of implementation.

During the workshop, the researcher gained practical experience in system design approaches. These approaches are useful for analysing how new scientific methods and digital tools are introduced into institutional and regulatory systems, what barriers may limit their uptake, and how tools should be designed so that they are not only technically valid, but also usable in real decision-making settings.

This is directly relevant to the later stages of the project, where a dynamic decision-support tool will be developed, validated and tested with potential users. The tool will need to integrate data and models, but it will also need to provide outputs that are understandable, transparent and useful for risk assessors, laboratories, competent authorities and other stakeholders.

During the CHANGE workshop, new potential cooperation partners were identified and contacted regarding possible future collaboration. These contacts are particularly relevant for work on New Approach Methodologies, implementation barriers, system design and practical use of decision-support tools in risk assessment.

Development of international cooperation networks

During the reporting period, international cooperation was further strengthened through participation in COST Action CA22105 BeSafeBeeHoney and through the identification of new cooperation opportunities relevant to later stages of the postdoctoral project.

In addition to the ongoing work within BeSafeBeeHoney, the newly approved COST Action CA25159 “From Data to Decision in Transforming Agrifood-Systems – Operationalizing One Health through AI”, also known as Agrifood-OneAI, was identified as a highly relevant cooperation platform for the later stages of the project. COST lists Agrifood-OneAI as running from 22 September 2026 to 21 September 2030. Its scope includes the use of artificial intelligence to operationalise One Health in agrifood systems, with attention to decision-making, data interoperability and risk management.

This COST Action is closely aligned with the project’s “from data to decisions” direction. Its focus on artificial intelligence, data infrastructures, One Health and decision support can provide useful links for the project’s later work packages, particularly those related to dynamic tool development, validation, user engagement and dissemination.

Scientific outputs, publications and knowledge transfer

During the reporting period, work continued on scientific publication concepts linked to the project. One of the main directions is a review article on risk classification and prioritisation methods in food safety, building on the methodological work carried out during the Maribor mobility and the Risk Thermometer analysis.

Work also continued on the structuring of occurrence data from scientific literature. This direction supports the development of the Occurrence Router component and addresses a recurring problem in food safety risk assessment: important occurrence data are often available in scientific publications, but not in a format that can be directly used for modelling, prioritisation or decision support.

In parallel, the researcher continued academic and knowledge transfer activities, including consultation of student work and lecture or seminar activities. These activities support the transfer of project-related expertise in data science, risk assessment and chemical food safety to students and early-career specialists.

Public communication and next steps

Public communication and science outreach activities also continued during the reporting period. A popular-science article on food safety topics is under preparation. Participation in public science communication activities is also planned, including the “Mērsraga ābols” event in August 2026 and activities related to European Researchers’ Night in September 2026.

In the next project period, work will continue on the integration of health-based guidance values and predictive toxicology data, the use of food consumption data, and the preparation of exposure assessment models. Further development of the Occurrence Router component and adaptation of the Risk Thermometer methodology will also continue.

The research application is implemented within the European Union Cohesion Policy Programme 2021–2027, Specific Objective 1.1.1 “Strengthening research and innovation capacity and introducing advanced technologies into the common R&D system”, measure 1.1.1.9 “Postdoctoral research”, under agreement No. 1.1.1.9/1/24/I/001. The research application is co-financed by the European Regional Development Fund.

The total approved costs of the project are in the amount of 184,140.00 EUR, of which 156,519.00 EUR or 85.0% is European Regional Development Fund financing and 18,414.00 EUR or 10.0% is state budget financing.

The project implementation duration is 36 months – from July 1, 2025, to June 30, 2028.

30.06.2026.