← 5ChAGI
Not yet open. Prices and scope below are indicative while the intake
and consent process is being finished. Nothing is being sold and no file should be sent yet.
Ancestry analysis
A formal admixture analysis of your consumer genotype file, fitted against
published ancient DNA with the same models and the same refusals used in the published research.
You send a 23andMe, AncestryDNA or MyHeritage raw file. You receive a dossier:
ancestry proportions with intervals at two model depths, your position among regional reference
populations, your paternal and maternal lines, and the receipts to reproduce all of it.
What makes this different
Most consumer ancestry results are a pie chart with no error bars and no way to check them.
This is the same method used in academic work — qpAdm over the Allen Ancient DNA
Resource — and it ships with its own limits written down.
- Intervals, not decimals. A share of 43 ± 3% is reported as 43 ± 3%. A single decimal
with no interval is precision nobody has.
- A reference set judged by what it rejects. Ours correctly refuses ancient sources
that cannot have contributed, and removing any one of its twelve members destroys those
refusals. A reference set that accepts everything has measured nothing.
- Receipts. Model definitions, weights, standard errors and software versions are
included so another analyst can rerun it.
Pricing
Dossier
$180
- Three-source admixture model with intervals
- Placement against regional reference populations
- Paternal (if male) and maternal haplogroup
- Written receipts and methods
Extended dossier
$320
- Everything in the dossier
- Deeper four-source decomposition
- Unsupervised clustering across K = 6–12
- Comparative f-statistics against named ancient populations
- A written answer to one question of your choosing, or an honest statement that the
data cannot answer it
Additional file
$90
- A second family member analysed under the identical model
- Ordered alongside either tier
Prices in USD. Each dossier includes a human review
pass before delivery — that review is what makes the result defensible, and it is why this is
not a volume product.
What the analysis will not claim
This is not a health or trait report. No health-associated variant in your file is
examined, stored, or reported. If you want medical interpretation, this is the wrong service.
No caste, clan, tribe or village of origin. Genetics does not carry those labels.
Anyone selling them is selling inference as fact.
No shared-segment "match" to a named ancient individual. That method needs
imputation and a reference panel not applied to consumer files, and a small apparent match
sits below the validated floor of the method. It is a lead, never evidence.
No fit statistic sold as proof. A single genome is statistically easy to fit — the
same models are rejected outright when applied to well-sampled populations. You are sold
proportions and placement, not a certificate.
How it works
- Ask. Email with your file type and what you want to know. You get a scope and a
price before anything is sent.
- Consent, then upload. Payment and a written consent come first; a private,
single-use upload link follows. Do not email raw genotype files — they will not be
accepted that way.
- Analysis. Your file is merged to the ancient panel and the models are run and
reviewed by hand.
- Delivery. You receive the dossier and the receipts.
- Deletion. Your raw file and every intermediate are deleted on delivery unless you
ask in writing that they be kept for follow-up work. You can request deletion at any time
and get written confirmation.
Your data
- Your file is processed for your analysis only. It is not used to train anything, resold,
or shared with a third party.
- Uploads are encrypted in transit and at rest, and raw files never appear in a log.
- Health-associated variants are never examined or reported.
- Deletion on request, confirmed in writing.
Enquiries
Scope, questions, or a request to be told when this opens:
[email protected].
Say which file you have and what you are hoping to learn.