Method, evidence and limits
We turn a connected family history into an individual estimate of carrier and onset risk, with the uncertainty stated rather than hidden.
At a glance
Family history is evidence. We treat it as a connected whole, not a checklist of symptoms or a textbook fraction.
12,867
conditions represented
11,068
with a documented monogenic inheritance pattern
Two-part review
published, molecularly confirmed families and controlled simulations
Our purpose
Questions about inherited conditions are personal, and generic percentages rarely answer what a family actually needs to decide. What You Inherit was built to make the reasoning behind genetic risk inspectable — before, during or after a conversation with a qualified professional.
The service estimates the chance that a person carries a relevant disease-associated variant, and the chance they may develop the condition given current age and penetrance. It is designed for education and personal understanding, with language that explains uncertainty rather than concealing it.
How the estimate is formed
The estimate considers the family as a connected structure. Who is affected, who remains well, how people are related, current age, age at first symptoms and the pattern of symptoms all contribute evidence inside one framework.
The same fact can carry different weight in different families. An unaffected person at 70 is usually more informative than an unaffected person at 20. A condition appearing across successive generations points in a different direction from one appearing only among siblings.
Possible explanations are compared against the complete family record. The result is a ranked estimate with explicit uncertainty — not a deterministic answer and not a replacement for laboratory testing.
Evidence and review
The disease catalogue draws on large-scale sources covering human genes, inherited conditions, genetic variation and documented modes of inheritance. Curated material is reviewed as the scientific record develops.
Performance is checked in two complementary ways: against published families whose diagnosis was later confirmed genetically, and against simulated families where the underlying answer is known in advance. Checks are repeated when the model or its source material changes.
Important distinction: benchmark performance describes how the method behaves on selected evaluation material. It is not a claim of clinical accuracy for every person, family, ancestry or condition.
Our principles
- Clarity over certaintyWe present an estimate as an estimate and explain what can change it.
- The family in contextRelationships, age and unaffected relatives matter alongside symptoms.
- Restraint by designWe do not turn a probability into a diagnosis or a treatment recommendation.
- Room for professional judgementResults are intended to support better questions for a doctor or genetic counsellor.
Scope and limitations
Every estimate depends on the accuracy and completeness of the information entered, and on how well available evidence represents the family concerned. Incomplete histories, uncertain diagnoses, small datasets, ancestry differences and conditions with complex or unknown causes can all widen uncertainty.
What You Inherit is not a diagnostic service, does not provide medical advice and does not establish a clinician–patient relationship. Decisions about testing, treatment, pregnancy, insurance or finances should be made with an appropriately qualified professional.
Data sources and attribution
The estimates rest on public scientific resources. Each is used under its own terms, credited here, and none of them endorses this service.
- Human Phenotype Ontology (HPO), annotation release 2026-09-02 — the vocabulary of symptoms and each condition's features. See hpo.jax.org.
- Orphanet (Orphadata, release 2026-06-23; CC BY 4.0) — prevalence, inheritance, age of onset and gene links of rare diseases. Orphadata: Free access data from Orphanet. © INSERM 1999. Available at orphadata.com. Figures are recombined, not reproduced.
- Mondo Disease Ontology (release 2026-09-01; CC BY 4.0) — which OMIM and Orphanet records describe the same disorder.
- GenCC (CC0) — the strength of each gene–disease relationship, from ClinGen, PanelApp, Orphanet and others.
- ClinVar and GeneReviews, resources of the National Center for Biotechnology Information (NCBI), U.S. National Library of Medicine — pathogenic variants, and published facts such as penetrance and detection rates, each cited in the source tables. NCBI does not endorse this service and is not responsible for how the data are interpreted here.
- Genome Aggregation Database (gnomAD v4, via Ensembl VEP; CC0) — how common each pathogenic variant is, by population.
- SEER Program, National Cancer Institute (SEER*Explorer, April 2026 release; public domain) — population cancer incidence by age and sex.
- ClinGen Actionability, Clinical Genome Resource (adult and pediatric curations, accessed September 2026; CC0), actionability.clinicalgenome.org — what can be done about a diagnosis: the outcome at risk, the intervention and ClinGen's scores. ClinGen's curations are not for diagnosis or medical decisions without review by a genetics professional.
- Medical Action Ontology annotations (MAxO; Carmody et al., Med 2023; CC BY 4.0) — medical actions that treat or prevent a disorder's features. Used as published, filtered to four relations.
- Fetal Medicine Foundation — the chance of trisomy 21, 18 and 13 by the mother's age and the week of pregnancy (Snijders and Nicolaides, The 11–13+6 weeks scan, 2004; Snijders et al., 1995 and 1999).
- Published cohort studies and national screening studies for penetrance, prevalence, new-mutation shares, repeat-size classes, Chinese carrier frequencies and carriers' cancer risks, each cited row by row.
Contact
For questions about the service, its methodology, an account or a published correction, email whatyouinherit@gmail.com. We review substantive feedback and use it to improve the clarity and quality of the service.
Head office: What You Inherit, Inc., 160 Spear Street, Suite 1300, San Francisco, CA 94105, United States.