Evagene changelog
Release notes for Evagene's pedigree platform. Dated entries, with links to the full write-up for each release.
New — a single-page complete feature list
A new Features page collects everything Evagene does in one place: gesture pedigree drawing and NSGC notation, the guided family-history intake questionnaire, the 230+ disease catalogue, the twenty-one published risk-model algorithms across four families, HPO phenotyping and phenotype-to-gene, the related-concepts graph, consanguinity and IBD, AI-assisted draft summaries, the eleven import / export formats (GEDCOM, GA4GH Phenopackets v2, CanRisk, 23andMe and more), and the full platform layer — REST API, BYOK LLM, webhooks, analysis templates, the 27-tool MCP server, and the embeddable viewer. It also indexes the free, MIT-licensed repository of 25 open-source integration examples at github.com/xiasma/evagene-integration-examples. No new product capability — a documentation page, consistent with Evagene's positioning as a research and education tool.
Genetics expansion — HPO phenotyping, Phenopackets, relatedness, and more
Six new educational and research features. HPO phenotyping lets users browse and search a curated Human Phenotype Ontology catalogue and attach observed or excluded phenotype features (with onset and severity) to individuals. Phenotype-to-gene educational associations surface candidate associated genes from the recorded HPO features — reference data only, never a ranked likelihood, a diagnosis, or a recommendation. The 230+ disease catalogue now carries MONDO, Orphanet, and Disease Ontology (DOID) cross-references alongside ICD-10 and OMIM. GA4GH Phenopackets v2 joins GEDCOM 5.5.1 and CanRisk / BOADICEA as an import and export format. A new relatedness / IBD module (gated for educational and research use) computes pairwise kinship, per-individual inbreeding, Jacquard k0 / k1 / k2 identity coefficients, and a gene-dropping relatedness simulation — never a forensic tool. And an illustrative educational prostate-cancer family-history risk model brings the total to twenty-one published risk-model algorithms; where it cites a published threshold it attributes it to the external guideline and states that Evagene does not make that recommendation. The MCP server grows to 27 tools. All of this is educational and research reference data and documentation: HPO phenotyping and phenotype-to-gene produce no risk number, nothing is diagnosed or asserted about any person, and Evagene remains not a medical device.
Family history questionnaire — guided intake
A new guided, step-by-step questionnaire records a family and personal health history and turns it into a pedigree. Answer questions about the subject and their relatives — living status, ages, conditions, and test results — and Evagene creates the individuals, wires up the relationships, sets the proband, and records the findings, adding any connecting relatives automatically. It is reached from the dock (“Family history intake”) or from the Pedigree builder panel. This is educational data capture, not a medical assessment: it runs no risk model and produces no risk number. See the user-guide page.
Related Concepts — an educational correlation graph
A new way to explore how Evagene's catalogues connect. The correlation graph is a hand-curated set of educational associations (over 1,100 curated edges) between diseases, traits, clinical-test results, allergies, and markers, plus genetic edges (genetic_association, shared_gene) derived automatically from the marker catalogue. A new Related concepts panel in the editor surfaces, per recorded item, the concepts associated with what is on an individual — with status-qualified clinical-test edges (a low ferritin and iron-deficiency anaemia; a high ferritin and hereditary haemochromatosis). The same data is available through the open /api/correlations* endpoints and a new related_concepts MCP tool (the fifteenth on the server). It is educational reference data: it does not diagnose, produces no risk number, and asserts nothing about any person — nothing is recorded by inference.
Evagene Pedigree Builder — ChatGPT Custom GPT
A new natural-language route into Evagene. The Evagene Pedigree Builder is a published ChatGPT Custom GPT that converts a plain-English family description into an Evagene-compatible pedigree JSON file. The user describes their family, the GPT asks clarifying questions where structure is ambiguous, and it returns a downloadable file that imports directly into Evagene via File → Load pedigree → Load Evagene file. This is a Route A integration (JSON-producing); a Route B with live Actions against the Evagene API is planned but not yet built. Requires a paid ChatGPT subscription — OpenAI's policy for all Custom GPTs, not an Evagene paywall. The GPT is a data-capture aid only: it does not give medical advice, diagnose, or run risk models; all interpretation happens inside Evagene after the file is loaded.
Four non-classical inheritance models
X-linked dominant with five sex-differential severity sub-modes (Rett-class, incontinentia pigmenti, CFND, EFMR); mitochondrial with strict maternal transmission and heteroplasmy scaling (LHON, MELAS, MERRF, NARP, Leigh, Kearns-Sayre, Pearson); digenic two-locus with the classical 25% ratio (Usher type 2, some RP, primary congenital glaucoma); imprinting / uniparental disomy with mechanism-weighted recurrence and the parent-of-origin rule for IC defects (Prader-Willi, Angelman, Beckwith-Wiedemann, Silver-Russell, TNDM).
Polygenic inheritance, nine cancer family-history models, CanRisk export
Liability-threshold recurrence-risk engine (Carter 1961, Falconer 1965, Reich/James/Morton 1972) across 20+ complex diseases. Nine cancer family-history scoring models — Claus, Couch, Frank, Manchester, NICE CG164/NG101, Amsterdam II, revised Bethesda, Gail NCI BCRAT, and a Tyrer-Cuzick IBIS-style approximation. One-click CanRisk / BOADICEA v4 pedigree export (BOADICEA not bundled — licensed by the University of Cambridge). Reproductive / breast-health fields added to the Individual record to drive Gail, Tyrer-Cuzick, and the CanRisk export.
Also see
- Research & citations — peer-reviewed papers citing Evagene and the U.S. VA Technical Reference Model listing.
- Decision matrix — which risk model for which counselling question?
- Complex-disease pedigree software — the 20+ catalogued multifactorial conditions.
- Mitochondrial inheritance · Imprinting / UPD · X-linked dominant · Digenic