Families living with rare monogenic epilepsies face profound challenges. For many of these conditions, scientific understanding and treatment options remain limited. Each condition is individually rare, patient populations are small, and the information that could drive discovery is spread across many institutions, registries, and advocacy groups. Often the data needed to make progress exist, but are difficult to find and connect.

With support from the Chan Zuckerberg Initiative, Data for the Common Good is working alongside the rare epilepsy community to change that, connecting existing efforts and building shared resources that can accelerate research and care.

Our approach: building together

Breakout discussion at CHIMES

Breakout session at CHIMES

The rare epilepsy community has already built many resources: registries, advocacy networks, datasets, and platforms, most created around one specific condition or gene. Our aim is to help these efforts connect and to build resources that can be shared. Though individual diseases differ, the underlying needs for research progress are common: ways to convene stakeholders, agree on data standards, and make existing data easier to find and use.

Our work started with listening. More than 70 clinicians, scientists, patient advocates, and caregivers came together at our CHIMES summit in 2025. One specific request from this community was a clear picture of what already exists and how it is connected. That picture is the Living Atlas.

The Living Atlas of Rare Epilepsies

The Living Atlas is an interactive map of the people, groups, and data that make up the rare epilepsy world—genes, conditions, advocacy and family organizations, registries, and the platforms that hold data—and how all these pieces relate to one another. The goal is a shared, clear picture of what exists, so families, researchers, and advocates can see who is working on what, where efforts connect, and where the gaps remain.

Screenshot from the Living Atlas showing the reference card and network map for the gene KCNT1

With the Atlas, you can:

  • Explore by gene or by condition and watch the connected community take shape around it.
  • See advocacy and family groups, and how they relate to and overlap with one another.
  • Look at any entry three ways: a reference card with links to trusted sources, a visual network map, and an AI-assisted chat that answers questions from those same sources.

It’s designed as a living resource rather than a static report: searchable, structured, and continually updated and corrected as the community grows.

Coming soon.

A data commons for rare epilepsies

The next step is a publicly accessible, integrated home for monogenic epilepsy data: a data commons built on voluntary participation, where those who contribute help set the rules. 

Combining expertise from the community with D4CG’s experience building data commons for rare diseases, we aim to bring together clinical, genomic, EEG, and other data from medical centers, registries, and patient advocacy groups, and to harmonize the data to one standard so that information from different sources can be used together for research. This work will require forming a consortium and shared governance, developing a common data dictionary, and establishing infrastructure to search and explore what is available. This effort is in its planning stages, with Citizen Health as the first data contributor.

Ethical AI for rare disease research

By incorporating community engagement and ethics throughout the entire development of a project, multimodal artificial intelligence (MAI) can be more equitably implemented, trusted by the community, and responsive to the community’s needs. With a team of University of Chicago researchers and other stakeholders, we are working on an MAI approach for monogenic epilepsies with the principles of responsible development and community involvement built directly into the work, developing an ethical framework for further MAI research in pediatric rare diseases. 

Graphic showing MAI ethics engagement steps for Feature Selection and Model Design, Data Collection, Training and Evaluation, and Deployment.

MAI Ethics engagement steps

Led by Alan Liang, Dr. Wim van Drongelen, and Dr. Douglas Nordli, the team’s approach uses AI to draw on information from EEGs together with clinical and genetic data to explore what these signals can reveal about brain function. Community engagement and ethical review are part of the process at every stage, led by a stakeholder engagement panel bringing together the perspectives of families, clinicians, and others close to these conditions. The panel reviews the model’s results and weighs their ethical implications, and that input feeds back into how the model is refined. The underlying research, which will soon be shared on public research repository medRxiv, points toward noninvasive ways to understand monogenic epilepsies and, in time, to assess how new therapies are working.

Stay in touch!

We look forward to sharing our progress. Join our email list to receive occasional updates!