pedigreejs vs Evagene: an open-source pedigree editor vs a full pedigree platform

Short version. pedigreejs is a free, open-source (GPL) JavaScript and D3 web-based graphical pedigree editor, maintained by the CCGE / BOADICEA group at the University of Cambridge — the same team behind BOADICEA and the CanRisk ecosystem. It is a client-side library you embed in your own page: it draws pedigrees interactively in the browser, and it loads and exports CanRisk and PED files. It ships with no bundled risk models, no hosted API, no AI or LLM tooling, no MCP server and no Custom GPT; it is the editor component, and the risk computation lives elsewhere at canrisk.org. Evagene is an academic, research and educational pedigree modelling platform — a hosted application that adds 20 published risk-model algorithms, a 230+ disease catalogue, AI-assisted draft summaries for educational and research review, an MCP server, a ChatGPT Custom GPT, a REST API and an embeddable viewer. Both run in the browser and both export CanRisk pedigree files. If you are a developer who needs only an embeddable drawing widget and will build the rest yourself, pedigreejs is very likely the better, cheaper choice. If you want the surrounding platform without building it, that is what Evagene is for.

This article is an honest, fact-by-fact comparison. Where pedigreejs is the better fit, we say so plainly. All claims about pedigreejs are drawn from its public GitHub repository, its 2018 Bioinformatics paper, and the CanRisk project documentation as of June 2026; if anything has since changed, those sources supersede this article. Evagene is an academic, research and educational pedigree modelling platform and is not a medical device; its outputs are illustrative and for educational and research purposes only.

How the two projects position themselves

pedigreejs is, by its own description, a web-based interactive tool to draw pedigrees — an open-source JavaScript library built on D3. It is maintained by the Centre for Cancer Genetic Epidemiology (CCGE) at the University of Cambridge, the group responsible for BOADICEA and CanRisk, and it is the editor component used inside the CanRisk ecosystem to capture family structure before risk computation. Its scope is deliberately tight: render and edit a pedigree in the browser, and read and write the CanRisk and PED file formats. You include the library in your own page and wire it up yourself. There is no hosted service, no account system, no support contract and no subscription — it is code you self-host or integrate, under the GPL.

Evagene positions itself as an academic, research and educational pedigree modelling platform. The pedigree is still central — gesture and keyboard drawing, standard ISCN/HGNC notation, a family register and a karyogram — but around it sit a suite of 20 published risk-model algorithms, a 230+ disease catalogue coded to ICD-10 and OMIM, an educational inheritance-model layer, a Related Concepts correlation graph for teaching, AI-assisted draft summaries for educational and research review via bring-your-own-key LLMs, and a platform layer (REST API, embeddable viewer, MCP server, Custom GPT, Analysis Templates) that lets pedigree data flow into other tools and AI agents. It is delivered as a managed application, free during its Alpha via a waiting list.

The headline difference is one of layer, not of quality. pedigreejs is a focused, well-maintained drawing library that does one thing and does it cleanly. Evagene is a broad platform that includes pedigree drawing as one capability among many and wraps it in catalogues, published risk models, agent tooling and a managed service. A developer who wants a widget should read this as "library versus platform", not "worse versus better".

Feature-by-feature comparison

The matrix below is arranged by capability. A tick means the feature is publicly documented; a dash means it is not publicly documented (which does not necessarily mean it is absent). Because pedigreejs is a drawing library and Evagene is a platform, many rows will tick on one side and dash on the other — that is the point of the comparison, not a flaw. Where nuance matters, we add a note.

Matrix compiled from the pedigreejs GitHub repository, its 2018 Bioinformatics paper, and Evagene public product pages as of June 2026. "—" indicates the capability is not publicly documented and does not necessarily mean it is absent.

Pedigree drawing and embedding

Both tools draw pedigrees in the browser, and both can be embedded in someone else's page. pedigreejs is built precisely for that: it is a D3-based JavaScript library you drop into your own application, and the pedigree renders client-side with no server round-trip. For a developer who wants a clean, well-understood drawing component maintained by the Cambridge CanRisk team, that is an excellent fit — lightweight, self-contained, and free to modify under the GPL. If your requirement is "render and edit a pedigree inside my own app", pedigreejs is hard to beat on simplicity and cost.

Evagene also offers an embeddable pedigree viewer, but it sits at a different level: the viewer is a window onto pedigrees managed inside the hosted platform, rather than a library you compile into your build. Evagene's own drawing surface adds gesture and keyboard construction, a family register and a karyogram, and standardises symbols against ISCN/HGNC notation. The useful point for developers is that Evagene pairs that embeddable viewer with open agent tooling — an MCP server and a Custom GPT — that pedigreejs does not provide. So if your goal is purely an in-page drawing widget, pedigreejs is the leaner answer; if you also want an AI agent to be able to read or build the pedigree, Evagene's surfaces extend further.

CanRisk export: the genuine overlap

The clearest shared use-case is the CanRisk file. pedigreejs reads and writes CanRisk and PED files natively — unsurprisingly, since it is the editor component of the CanRisk ecosystem built by the same Cambridge group. Evagene exports a ##CanRisk 2.0 pedigree file that the user uploads at canrisk.org to run BOADICEA. BOADICEA is not bundled in Evagene; it is licensed by the University of Cambridge, and the CanRisk file is an off-platform bridge, not an on-platform computation. Evagene treats that architectural separation as a feature: the authoritative risk number is produced at canrisk.org, not inside Evagene.

So on the specific task of "produce a CanRisk file for BOADICEA", the two tools genuinely overlap, and pedigreejs — being the native CanRisk editor — is a perfectly good way to do it. The difference is everything around that file: pedigreejs hands you the editor and stops there, while Evagene wraps the same export in a catalogue, an inheritance-model layer, illustrative risk models for conditions outside BOADICEA's scope, and AI-assisted draft summaries for educational and research review.

Risk models and inheritance (educational and illustrative)

pedigreejs has no bundled risk models — by design. It captures the pedigree; the risk computation in the CanRisk ecosystem happens server-side at canrisk.org. That is a clean separation and entirely appropriate for a drawing library.

Evagene bundles 20 published risk-model algorithms as an educational and research layer, with outputs framed as illustrative rather than as clinical results. The set covers BayesMendel (BRCAPRO, MMRpro, PancPRO); Mendelian and non-classical inheritance (autosomal dominant, autosomal recessive, X-linked, mitochondrial, digenic and imprinting); a liability-threshold multifactorial / polygenic / oligogenic engine (Carter 1961, Falconer 1965); and cancer family-history scoring — Claus 1994, Couch 1997, Frank 2002, Manchester (Evans 2004), NICE, Amsterdam II (Vasen 1999), revised Bethesda (Umar 2004) and Gail 1989. It also includes a Tyrer-Cuzick model that is an IBIS-style approximation of the published Tyrer/Duffy/Cuzick 2004 algorithm, not the official IBIS Breast Cancer Risk Evaluator binary — for the official Tyrer-Cuzick computation you would use the IBIS tool itself, and for BOADICEA you would use canrisk.org. Every one of these outputs in Evagene is illustrative and for educational or research purposes only; none is a diagnosis, a recommendation, or a substitute for professional judgement.

AI, agent tooling, and the platform layer

This is where the two projects diverge most, and it is squarely a library-versus-platform difference. pedigreejs has no AI or LLM tooling, no hosted API, no MCP server and no Custom GPT; it is a client-side editor, and that is the whole intent.

Evagene treats agent and platform surfaces as first-class. It provides AI-assisted draft summaries for educational and research review using bring-your-own-key LLMs (Anthropic Claude, OpenAI) — your own model account drafts an educational summary that a human reviews; the output is illustrative, not medical advice. It ships an MCP (Model Context Protocol) server with 15 tools so AI agents such as Claude Desktop and Claude Code can read, build and analyse pedigrees; the "Evagene Pedigree Builder" ChatGPT Custom GPT, which turns natural-language family histories into importable pedigree JSON; a Related Concepts educational correlation graph (1,100+ curated associations — reference data for teaching, not risk analysis or diagnosis); a guided Family History Questionnaire that turns structured intake into a pedigree (educational data capture, not a medical assessment); Analysis Templates; and a scoped REST API with webhooks and the embeddable viewer. None of this is part of pedigreejs's remit, and none of it is a criticism of pedigreejs — it is simply a larger surface area than a drawing library sets out to provide.

Cost, licensing, and who maintains them

pedigreejs is free and open-source under the GPL, with the source on GitHub and a peer-reviewed 2018 Bioinformatics paper behind it. You self-host or integrate it; there is no SaaS, no subscription and no support contract, and you are free to read, fork and modify the code. It is maintained by the CCGE / BOADICEA group at the University of Cambridge — strong provenance for anyone already working in the CanRisk world.

Evagene is a hosted platform, free during its Alpha via a waiting list, with general-availability pricing not yet published. It is delivered as a managed service rather than as source you compile, so the trade-off is the usual one: you give up the control and zero-cost-forever properties of self-hosted open source in exchange for not having to build, host or maintain the catalogue, the risk models, the agent tooling and the API yourself. For a team without the appetite to build all of that, that trade can be worthwhile; for a developer who only needs the drawing widget, it is unnecessary overhead.

When to choose pedigreejs

• You are a developer who needs an embeddable pedigree drawing widget inside your own application, and you will build the rest of the stack yourself.
• You want free, open-source code under the GPL that you can read, fork and modify, with no SaaS dependency.
• Your main task is producing CanRisk or PED files, and a native CanRisk editor is exactly what you need.
• You value provenance from the Cambridge CCGE / BOADICEA team and want to stay close to the CanRisk ecosystem.
• You want a lightweight, client-side component with no account system, no server dependency and no ongoing cost.

When to choose Evagene

• You want the surrounding platform — disease catalogue, published risk models, agent tooling — as a managed service rather than something you build on top of a drawing library.
• You want 20 published risk-model algorithms and an educational inheritance layer alongside the pedigree, with outputs framed as illustrative for teaching and research.
• You want AI-assisted draft summaries for educational and research review using your own LLM keys, an MCP server for AI agents, and a Custom GPT for natural-language pedigree building.
• You need a 230+ disease catalogue (ICD-10, OMIM), a Related Concepts educational graph, and a guided Family History Questionnaire.
• You want broad import/export (GEDCOM, JSON, 23andMe, image-OCR; export PNG/SVG/PDF/DOCX/CanRisk) plus a REST API, webhooks and an embeddable viewer without building them.
• You still want a one-click CanRisk file for BOADICEA at canrisk.org — Evagene shares that use-case with pedigreejs and adds everything around it.

Can you use both together?

Yes, and for some teams that is the most natural answer. Because both speak CanRisk and PED, a pedigree edited in pedigreejs inside your own application can be exported as a CanRisk file and brought into Evagene's hosted platform — or the reverse — without re-entry. A developer might embed pedigreejs as the drawing component in a bespoke app while using Evagene's REST API and MCP server for the catalogue, the illustrative risk models and the AI-assisted educational summaries. The two are complementary rather than mutually exclusive: pedigreejs is the open, embeddable editor; Evagene is the managed platform around the pedigree. Both ultimately route serious BOADICEA risk computation off-platform to canrisk.org, where BOADICEA is licensed and hosted by the University of Cambridge.

Frequently asked questions

Related comparisons and reading

• CanRisk vs Evagene
• GenoPro vs Evagene
• QuickPed vs Evagene
• Compare Evagene with other pedigree tools
• Hereditary cancer risk assessment