⚡EpilepsyLive ✉ Suggest a correction
Demonstration project by Danielle Boyce. Not medical advice. Not a production system, and not affiliated with, endorsed by, or representing any patient advocacy group, registry, company, or institution named. Any participants and records shown anywhere on the site are fictional. Information is drawn from public sources, may be incomplete or out of date, and any errors are unintentional.

EpilepsyLive guide

What each page is for, and what every term on it means. If something here is unclear or wrong, use Suggest a correction at the top of any page.

Start hereWhat is EpilepsyLive

EpilepsyLive is a set of connected tools for the rare epilepsy research and advocacy community. It has four parts, switchable from the bar at the top of every page.

  • Live Monitor. A live feed of published research, preprints, and clinical trials for epilepsy genes, drawn straight from public databases. Nothing is stored; it queries the sources each time you change a filter.
  • Registry Toolkit demo. A walkthrough and planning aid for a community that is thinking about building a research registry. It shows what a participant would experience, and collects the reference material and planning tools in one place.
  • PAG Connections. A directory of the patient advocacy groups (PAGs) and research infrastructure behind each gene, plus a network view of how those groups and conditions relate.
  • This guide. Definitions for everything on the site.
These are demonstrations and reference tools, not a registry. No page collects, stores, or transmits real participant data. Where the site shows people or records, they are fictional. Descriptions of organizations and infrastructure are compiled from public sources and may be out of date; corrections are welcome.

PageLive Monitor

The Live Monitor answers one question: what is being published and tested, right now, for a given epilepsy gene or topic. It reads from open, keyless services directly in your browser.

Where the data come from

  • Publications and preprints: Europe PMC, which includes PubMed and MEDLINE records and adds preprints and open access full text.
  • Clinical trials: ClinicalTrials.gov.
  • Known drugs and target evidence on the gene cards: Open Targets.
  • Funders: Crossref.

The three tabs

  • Literature: articles and preprints matching your query.
  • Trials: clinical trials matching your query.
  • Insights: charts summarizing the loaded literature, including top authors, top genes, publication years, and a co-occurrence network (defined below).

Filters and controls

  • Keyword box: free text added to the query, for example a therapy, technique, or biomarker.
  • Popular search terms: quick topic chips, for example infantile spasms or Dravet syndrome. Selecting one or more filters the feed to those topics.
  • Publication type: all, peer reviewed only, or preprints only.
  • Open access only: restricts to freely readable full text.
  • Sort: newest, most cited, or most relevant. Most cited uses Europe PMC citation counts.
  • Time: how far back to look.
  • Export: the loaded list can be saved to CSV at any time.
A note on "Most cited." Sorting by citations surfaces the papers the whole field cites most. For a broad topic query, that can float very large, heavily cited reviews or statistics reports to the top even when your topic is only a minor mention in them, because those papers are among the most cited in all of biomedicine. If the top results look off topic, try "Most relevant" or "Newest," or narrow the query with a specific gene or keyword.

Gene tagging

The Live Monitor recognizes epilepsy gene symbols in titles and abstracts and tags each result with the genes it mentions, using the same gene list that drives PAG Connections. That is how the Insights charts can count genes and their co-occurrences.

DefinitionLive Monitor: the co-occurrence network

On the Insights tab, the network shows which genes tend to appear together in the same papers across the results you have loaded.

  • Each circle (node) is a gene. A larger circle means the gene appears in more of the loaded papers.
  • A line (edge) between two genes means they were named together in at least one of those papers. Genes that are frequently studied alongside each other pull together.

This network is built on the fly from whatever set of papers is currently loaded, so it changes as you change the query, filters, or time window. It is a picture of co-study, not of biology: two genes linked here are written about together, which may reflect shared mechanism, shared syndrome, or simply a review that covers both.

PageRegistry Toolkit demo

The Registry Toolkit is a guided mockup for a rare epilepsy community weighing whether and how to build a research registry. The left navigation is grouped so it is clear which parts are a participant-facing demo and which are reference material.

What a participant experiences demo

These pages show the participant side of a registry: the overall flow ("What the Participant Sees"), a set of sample surveys, and a demonstration of how health records and consents would work.

The sample surveys are previews, not an endorsed instrument set. They exist to show the format and the kinds of questions a registry might ask, drawn from published instruments. They are not the complete list of what a real registry would collect, and their presence here is not an endorsement of any particular instrument. The authoritative item text lives in the Comprehensive Dictionary and its cited sources.

Health records and consent demo

Illustrative flows for how a family could bring a child's records into a registry: a live connection to a patient portal (SMART on FHIR), a downloaded and re-uploaded record file (C-CDA), network exchange (TEFCA / individual access services), a HIPAA authorization, and the special case of EEG and imaging studies. None of these connects to a real system here.

Reference and education

The material behind the demo: the Comprehensive Dictionary (the full, documented set of variables with their sources), Data Sources and References, and short guides on meeting a registry vendor, using EHR data, and implementation science.

Planning tools

The maturity model and its self-assessment (below), a Data Strategy Self-Assessment that helps match a study design to a purpose, a strategy library, and a resource directory of networks and partners.

DefinitionThe maturity model

The maturity model, in the Registry Toolkit, describes a path rather than a snapshot. Every rare epilepsy organization moves, at its own pace, from an informal community toward sharing standardized data that outlives any single group. The model names the stages on that path so an organization can see where it stands, where it is ahead, and where it is behind.

It reads as five tracks, each with five stages:

  • Organizational Foundation: from an informal community to a sustainable organization.
  • Scientific and Community Engagement: from peer support to helping set the research agenda.
  • Ecosystem Integration: from working alone to being a recognized node others come to.
  • Data Infrastructure: from no structured data to data harmonized to common standards.
  • Data Governance and Sharing, the north star: from no governance to open, responsibly shared, FAIR data.

An organization's overall level is roughly the lowest stage it has reliably reached across the tracks, because data cannot be shared responsibly without governance beneath them. The model is populated by self-assessment, not assigned from outside. Click any cell in the model for the questions to ask and the kind of support that moves an organization forward.

PagePAG Connections

PAG Connections maps the patient advocacy groups and research infrastructure behind rare epilepsy genes. It has two views, switchable from the tabs at the top of the page.

Directory view

The Directory shows one gene at a time. Pick a gene from the left rail (or search it), and its card opens with two lenses:

  • Reference lens (left): depth on that one gene. Identifiers (HGNC, Entrez, Ensembl, OMIM), inheritance and phenotype, the community infrastructure that exists for it (foundation, registry, Simons Searchlight, CoRDS, therapeutics, FDA patient-focused meetings), the registry platforms in use, the organization's social channels, and canonical reference links.
  • Network lens (right): that gene's place among conditions and organizations, with the counts summarized (advocacy organizations, related conditions, direct links, and its community) and an Open in network button that jumps to the Network view centered on this gene.
Why can I only pick one gene here? The Directory is deliberately one gene at a time, so each card can go deep. It opens on SCN2A as a worked example; pick any gene from the list, which is alphabetical, to switch. To see many genes, conditions, and organizations together, use the Network view. The search box in the top bar is global: it finds any gene, condition, organization, or syndrome name (for example "Gould syndrome" finds COL4A1) and jumps you to the right place.

DefinitionPAG Connections: the relationship network

The Network view is a map of how genes, conditions, and advocacy organizations connect. Unlike the Live Monitor network (which is built from papers), this one is built from the advocacy and infrastructure data set: who runs what, and which conditions share the same groups.

How to read it

  • Nodes come in three shapes: a circle is a gene, a square is a condition, and a triangle is an advocacy organization.
  • An edge (line) connects things that are directly related: a gene to the organizations that serve it, a gene to the conditions it causes, and a condition to the organizations that cover it. Two genes sit near each other when they share a condition or an organization.
  • Color groups nodes by community (defined below).

Key terms in this view

Community
A cluster of nodes that are more connected to each other than to the rest of the graph, identified by community detection on the relationship data. Nodes in the same community share color. A community often corresponds to a family of related conditions and the groups that serve them. The number (for example "Community 13") is just a label, not a ranking.
Direct links / degree
How many other nodes a node connects to directly. A gene with more direct links sits at the center of more relationships.
Related conditions
Conditions that surface through the organizations a gene shares. If two conditions are served by the same group, they show up as related.
Advocacy organizations
The patient advocacy groups, foundations, and family networks linked to a gene in the network layer.

Interacting

  • Search for a gene, condition, or organization to center the graph on it.
  • Click a node to recenter on its neighborhood; drag to pan; scroll to zoom.
  • Browse the largest communities from the side list to explore clusters.
Coverage. The Directory covers more genes than the network layer does. A gene can have a full reference card but not yet appear in the network, because the relationship data cover a subset. Widening that set is a planned step, not a redesign, since the two data sets already join on the gene symbol.

DefinitionPAG and Registry markers

Each gene in the Directory carries two simple markers, in the list and at the top of its card: PAG and Registry. They answer two plain questions from observable signals.

  • PAG: is there a patient advocacy group, foundation, or family organization dedicated to this gene or condition?
  • Registry: is there a research registry that collects data on it, whether run by the organization, an academic group, or a shared platform?

A gene shows a marker when that thing is present, and "No PAG" or "No registry" when it is not. That is all the markers claim.

Deliberately not a maturity score. Earlier versions showed a 1 to 5 level, which read like a graded assessment. None of these organizations have completed a maturity self-assessment, so a graded number would be misleading. The presence-based PAG and Registry markers only report what can be observed from the outside. The real, self-reported picture of where an organization stands lives in the maturity self-assessment in the Registry Toolkit, which an organization fills in for itself.

"Implicit" on the Searchlight and CoRDS rows

On a gene's card, the Community infrastructure block lists whether its community can enroll in two shared registry platforms, Simons Searchlight and Sanford CoRDS. Those rows can say "Yes," "No," or "Implicit."

Implicit means the gene is not listed by name on that platform, but participants would still be captured under a broader umbrella category, for example a general developmental and epileptic encephalopathy or intellectual disability grouping, rather than a gene-specific cohort. In other words, they could likely take part, just not through a dedicated entry for their gene.

ReferenceGlossary

PAG
Patient advocacy group. A community-led organization, foundation, or family network for a condition.
Registry
An organized collection of data about people with a condition, gathered over time for research and care.
Natural history study
A study that follows people with a condition over time to describe how it changes without a specific intervention. Often the evidence base a future trial is designed against.
Contact registry
A maintained list of who the participants are and how to reach them, an early step before a research-grade registry.
Comprehensive Dictionary
In the Registry Toolkit, the full documented set of variables a registry could collect, with each variable's source instrument.
SMART on FHIR
A standard that lets an app connect to a patient portal and pull a person's records with their consent.
C-CDA
Consolidated Clinical Document Architecture. A standard file a patient can download from a portal and upload elsewhere.
TEFCA / IAS
A national framework for health information exchange; individual access services let a person retrieve their own records through it.
HIPAA authorization
A signed permission that lets a specific party use or share a person's protected health information for a stated purpose.
Simons Searchlight
A research program and registry for many neurodevelopmental genetic conditions.
CoRDS
Coordination of Rare Diseases at Sanford, a shared registry platform many rare conditions use.
Implicit (Searchlight / CoRDS)
On a gene card, means the gene is not listed by name on that platform, but participants would be captured under a broader umbrella category rather than a gene-specific cohort.
FDA PFDD
Patient-Focused Drug Development. FDA meetings and listening sessions that capture patient experience for a condition.
FAIR
Findable, Accessible, Interoperable, Reusable. Principles for data that others can discover and use.
DEE
Developmental and epileptic encephalopathy. A severe group of genetic epilepsies with developmental impact.
North star
In the maturity model, the track that everything else builds toward: responsible data governance and sharing.

MethodsHow the PAG Connections network was built

This section documents how the relationship map is constructed so the results are reproducible and the gaps are honest. It is written for a general audience but is precise about the steps.

1. Starting point: a gene list

The map begins from a curated list of monogenic epilepsy genes (the same list that drives the directory). Each gene has a reference card with identifiers and clinical descriptors. The map does not invent genes; it only draws relationships for genes already on the list.

2. Three kinds of node

Every item on the map is one of three types: a gene, a condition (the disorder a gene causes), or an organization (a patient advocacy group, foundation, or registry host). Each is a node with a stable identifier.

3. Where the relationships come from

Edges are drawn only from recorded facts, never inferred from similarity. For each gene we record, from public advocacy and registry sources, which organizations serve that gene and which conditions it is associated with. An edge is added when, and only when, that link is present in the underlying data. A gene is joined to a condition using the condition name, and to an organization using the organization name, so the same real-world entity is one node no matter how many genes point to it.

4. Genes connect to conditions, not only to organizations

An earlier version linked genes to organizations and conditions to organizations, but never linked a gene directly to its condition. That made a gene look connected to advocacy groups while its conditions floated separately. The current map adds a direct gene to condition edge for every recorded gene and condition pair. Two genes that cause, or contribute to, the same condition are now connected through that shared condition, which is what makes the map feel unified rather than a set of separate spokes.

5. Communities and layout

Nodes are grouped into communities for color only. A community is a cluster of nodes that are more connected to each other than to the rest of the map, which usually corresponds to a disease area or a shared foundation. The layout uses a simple force based arrangement so that connected nodes pull together and unrelated nodes drift apart. Color and position are visual aids; they carry no meaning beyond grouping.

6. Degrees and counts

Each node reports its degree, the number of direct edges it has. The counts shown on a gene card (advocacy organizations, related conditions, direct links) are computed from the final edge set, so the numbers on the card and the lines on the map always agree.

7. What is deliberately not on the map

Some genes have a complete reference card but no recorded advocacy organization and no shared condition. These genes are shown in the directory but are not placed on the relationship map, because there is nothing in the data to connect them to yet. This is a real absence in the source data, not an error, and their cards say so plainly. As advocacy or registry links are recorded for them, they join the map automatically with no redesign, because the join keys already exist.

8. What the map is not

The map is a picture of recorded advocacy and infrastructure relationships. It is not a measure of research funding, publication volume, or how active a community is, and it is not a clinical or genetic similarity network. For a literature based view of how genes co-occur in the research record, use the Live Monitor network instead.

ProvenanceData sources

Live Monitor reads live from Europe PMC, ClinicalTrials.gov, Open Targets, and Crossref, so it reflects those sources at the moment you query. PAG Connections is built on a compiled genetic epilepsy landscape data set: the gene reference layer draws on a curated epilepsy gene list, and the network layer draws on a mapped set of advocacy organizations and the conditions they share. The Registry Toolkit's dictionary cites the published instruments each variable comes from.

The advocacy and infrastructure descriptions are compiled from public sources and are a snapshot; organizations change, registries launch, and links move. That is exactly the kind of thing the correction link is for.

ScopePrivacy and scope

Nothing on the site is a live registry. The Registry Toolkit does not collect, store, or transmit real data; its participants and records are fictional and its connections are demonstrations. The Live Monitor and PAG Connections query public databases and public compiled data and do not ask for anything about you. Any personal or clinical details you might imagine entering into a demo form stay in your browser and are not sent anywhere.

Help improve itFeedback

If a definition is unclear, a link is stale, or an organization is described wrongly or missing, please say so. Every page has a Suggest a correction link in the top bar, or write directly to danielle@boycedatascience.com. Corrections to the advocacy data are especially welcome, since that information ages quickly.

Something look wrong or unclear? Suggest a correction or send a note. EpilepsyLive is a demonstration project by Danielle Boyce, boycedatascience.com.