Conference: AI-Ready Data for Engineering Biology: Technical Standards and Quality for Accelerating Discovery and Innovation

Recent innovation in engineering biology has been driven by advances in biological artificial intelligence (AI) models that enable synthesis of large amounts of data to identify emergent trends and enable novel discoveries. Beyond simply increasing the volume of biological data (biodata) available to train new AI models, the crucial next step is the creation of AI-ready datasets that are standardized, well-annotated, and optimally formatted for advanced model training. However, the definition of AI-ready biodata, the parameters that define it, and the best practices for constructing such datasets are not well-established and vary across stakeholders and experts. Resolving these challenges will require building community consensus through structured stakeholder engagement of AI developers and engineering biologists, providing the critical foundation for future coordination, data standardization, and best practices development. The objectives of this workshop are to develop standards and best practices for building and curating AI-ready datasets. This project advances NSF's priorities in Artificial Intelligence and Biotechnology. Through this award, the Engineering Biology Research Consortium (EBRC) will convene a series of virtual stakeholder alignment meetings to define priority topics for an in-person workshop aimed at developing standards and best practices for AI-ready biodata. Over the course of several weeks, the stakeholder meetings will help ensure that a strong foundational framework is identified. Leveraging this initial work, the two-day in-person workshop will feature a combination of expert presentations, panel discussions, and small-group breakout sessions to stimulate discussions and refine the foundational framework. A key outcome of this work will be the development of a refined, community-led framework of standards, metrics and best practices for developing AI-ready biodata. In coordination with co-hosts from the National Institute of Standards and Technology, EBRC will disseminate results of the workshop through community education and outreach and follow-on activities. The framework will enable data generators and curators to ensure that their data is maximally reusable and interoperable, which will help accelerate innovation at the intersection of biology and AI. This work will also be particularly useful as different government agencies begin major data consolidation and curation projects. In short, this framework will enhance the quality of biodata sets, which in turn will drive progress and promote US competitiveness in biotechnology. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.