Challenge: In the drug development process, knowledge of the applied (pre)clinical models and biological endpoints, combined with their eventual authorisation, can contribute to understanding the success factors of the process, including possibly the identification of translational models. A potentially valuable data source for gathering this type of information is the European Medicines Agency (EMA), which makes various documents available via its website. To efficiently understand the strategies used in drug research, automating data extraction through text-mining techniques is essential.

Objective: This project aims to investigate whether text-mining can be successfully employed to extract knowledge from EMA documents. The focus is on drug-specific information regarding the applied (pre)clinical models, biological endpoints, and species. To determine this, a feasibility study will be conducted with a limited number of EMA documents—covering a specific indication and a small number of drugs.

The final result of this feasibility study will be a report containing:

• An analysis of the performance of the text-mining algorithm in extracting relevant data from EMA documents. Possible improvements in the used ontologies and strategies will also be reported.
• A potential approach for further developing EMA data analysis into a tool.

Project Status: This project started in mid-February 2023 and completed 2024.

Executor(s): TNO is carrying out this feasibility study.

Role of FAST: FAST was the sponsor of this feasibility study, advises the project team, and involves stakeholders in its execution.