Inventory of BMR values for BMD analysis

This report focuses on creating a repository of Benchmark Response (BMR) values sourced from international databases. The objectives included defining BMRs for specific endpoints and considering various characteristics such as compound characteristics, information on the study (e.g. strains, species and gender), information related to toxicity (target organ, type and mechanism of toxicity) and statistical aspects of the dose‐response modelling. The information was retrieved from repositories of risk assessment bodies such as the IRIS‐EPA, JECFA, and OpenFoodTox databases. The methodology section details data acquisition, sources pipeline, and software utilized, ensuring transparency and reproducibility. Techniques like web scraping and text mining complemented manual screening to retrieve and interpret JECFA assessments. OpenFoodTox data served as the reference, involving the filtering for Benchmark Dose (BMD) endpoints and mapping variables for a harmonized setting. When filtering IRIS‐EPA data, the code selected variables based on mapping information and filtered data from multiple sheets, aligning it with the OpenFoodTox reference. Moreover, the post‐processing steps are explained thoroughly, including data import and cleaning. The inventory results in 593 records in a wide format and offers valuable information on BMR values for BMD analysis, facilitating informed decision‐making for relevant future applications.