Advancements in Data Harmonisaton through Co-creation and AI-ready Tools: Insights and Future Outlook
Wednesday, November 13, 2024
3:00 PM to 4:30 PM (US EST, UTC -5)
Hosted by: The SOT Risk Assessment Specialty Section
Registration required for this free webinar.
Integration of computational methods and data-driven approaches is critical for advancing safety assessments and material design. Ideaconsult Ltd. has developed and maintained several open-source software solutions to facilitate these objectives. Dr. Jeliazkova will present the integration and harmonization of diverse datasets based on a common data model, leading to the establishment of Europe’s largest nanomaterial safety database, accessible at https://enanomapper.adma.ai. Key tools like the Template Wizard enable efficient capture and linking of experimental data and metadata using community-standard templates, making data generated from various instruments interoperable. While templates and the tools evolved with insight and experience gathered over a decade of EU funded projects addressing nanoscale materials safety, the templates are generalizable across fields and have already been extended and adapted for microplastics and advanced materials research. Automation of the protocol of creating new and modification of existing templates is established through user friendly Template Designer. ToxFairy, a High Throughput data preprocessing and analysis workflow is developed as open-source Python package and Orange add-on and communicates with the database via Application Programming Interface and harmonized formats. By making raw and processed data machine-readable, these tools enhance data reusability for automated analysis workflows and streamline the curation process, facilitating re-use across scientific contexts. Additionally, they introduce a natural language tool leveraging recent AI models to interact with the Adverse Outcome Pathway (AOP) knowledge base.
Speaker:
- Nina Jeliazkova, PhD, Ideaconsult Ltd.
Registration required for this free webinar.