One careful re-analysis. Yours to keep.
Drop in the genetic file you already have. We detect whether it is a genotyping-chip export, whole-genome data, or aligned reads, then show you what your report will cover before you pay. We re-read it and attribute every finding to its source, so you can check where a result came from. Every report stays yours for good.
Choose your starting point
Not sure which file you have? Drop it in and we will tell you. Otherwise, pick below.
Most whole-genome reports on GRCh38 are ready 5–15 minutes after the upload finishes. A GRCh37 file is lifted to GRCh38 first, and a large gVCF on that build can take a couple of hours. Deep Read takes about half an hour once the reads are in.
Chip report
For 23andMe, AncestryDNA, MyHeritage, FamilyTreeDNA, LivingDNA and similar genotyping-array raw-data exports. Everything your chip can call, fully attributed. Carrier status and the full clinical-findings panel need whole-genome data (WGS). MyHeritage kits sold since October 2025 may be sequenced rather than typed on an array, so check which file you have.
- Clinical findings: two clotting variants, Factor V Leiden and prothrombin G20210A
- Pharmacogenomics for the genes your chip typed and resolved
- Traits, ancestry and the Frontier module
- Polygenic scores and single-gene findings
- A one-page clinician hand-off, built from the same findings
- Plain-language explainers, save as PDF
Already on a chip report? Upgrade to the full WGS report for $20 any time. Chip plus upgrade comes to $39. The full report on its own is $39.
Full report (WGS)
Every module over your whole-genome file. Upload a gVCF for coverage-proven findings (a plain VCF also works). A one-time analysis, yours to keep.
- Clinical findings (ClinVar and ACMG SF)
- Carrier status and pharmacogenomics: 9 pharmacogenes from a plain VCF, 25 from a gVCF, each card stating how many defining positions your file examined
- The wider ClinVar screen: listed variants in 1,363 more recessive and X-linked genes, and 52 dominant genes rated actionable
- Traits, ancestry, Frontier and the Fringe
- 17 disease polygenic scores, with your percentile against a reference population, and 9 emerging-research scores kept apart from them
- A one-page clinician hand-off, built from the same findings
- Pinned, attributed, reproducible sources
Couple pack
Two genome reports, $78 separately. One for you and one for your partner, with the partner view built from the two.
Get the couple pack · $59Full report + Deep Read
Deep Read works out to $30 here, vs $49 on its own. The full WGS report plus the Deep Read panel, prepaid together.
- Everything in the full report
- Deep Read: CYP2D6 copy number and star alleles, from your reads
- Per-gene callability, measured directly
- Two recurrent deletions called from depth; HLA-DQ typed
Add when you are ready
Deep Read
The answers only your aligned reads hold: the BAM or CRAM your sequencing provider gave you, on GRCh38 or GRCh37. A variant file cannot see a gene's copy number, cannot say how much of a gene it actually examined, and cannot type HLA. Deep Read reads all of that from the reads, resolves the full 29-target pharmacogene panel to CPIC star alleles on the way, and turns it into a one-page hand-off for your prescriber. It costs a little more than the full report because it reads tens of gigabytes of aligned data with lab-grade star-allele calling. Most people save by taking it in the bundle.
- CYP2D6 copy number and star alleles: the duplication a variant file cannot see
- Per-gene callability, measured directly from read depth
- Two recurrent deletions (TYROBP, CLN3) called from depth
- HLA-DQA1 and HLA-DQB1 typed from the reads, which is what a celiac rule-out needs
- G6PD read as a single copy where that is what you carry
- The full 29-target panel (CYP2C19, TPMT, DPYD, SLCO1B1 and more) and a one-page medication hand-off
- Reads processed, then deleted immediately
Why does this panel need aligned reads?
CYP2D6 is the clearest example. It governs how the body metabolises dozens of common drugs, from antidepressants to codeine, and its gene structure (duplications, deletions, pseudogene interference) means a variant file cannot see whether you carry one copy or three. Most of the rest of the panel can now be resolved from a gVCF inside the full report, so what stays with the reads is exactly what needs depth: copy number, how much of each gene was actually examined, the recurrent deletions, and HLA typing. Deep Read measures, gene by gene, the fraction of bases we could examine at read depth of 10× or more, turning "not found" into "examined, and not found".
Offered from your dashboard once your report is ready, so there is no large upload up front. Prepay it with the bundle to save $19.
Start with the full report · $39Re-analysis subscription
$1 a month, billed once a year.
Your genome doesn't change. The science does. Keep a whole-genome report live for $12/yr: after we install new reference data, such as a ClinVar release, we re-read your variant file and your dashboard shows what changed. Without it, your report stays the snapshot you bought.
- Silent re-matching of your variant file
- "What changed" updates after each re-read, on no fixed schedule
- Only your normalised variants are kept
- You keep every report you have ever had. The subscription only adds new updates
- Cancel any time
Add it before your whole-genome report or within 30 days after it, while we still hold the variant file it re-reads. A chip report keeps no variant file, so the subscription does not apply to one. How updates work, and why a genetic result can change.
No hidden fees. No data resale (that is how the "free" services pay). One transparent price for a careful, attributed re-analysis you own.
What each file type can tell you
The same coverage table we show you before you pay. A number is how many items we examine in a module; a tick means included; a dash means that file type cannot deliver it.
| What you get | Chip | WGS | WGS + Deep Read |
|---|---|---|---|
| Clinical findings | Up to 2 | Up to 39 | Up to 39 |
| Carrier status | Not available | Up to 60 | Up to 60 |
| Wider ClinVar screen (known variants) | Not available | Up to 1,415[1], see note 1 | Up to 1,415[1], see note 1 |
| Pharmacogenomics | Up to 7[2], see note 2 | Up to 9[3], see note 3 | Up to 9[3], see note 3 |
| Traits | Up to 24[4], see note 4 | Up to 28 | Up to 28 |
| Frontier | Up to 56[4], see note 4 | Up to 60 | Up to 60 |
| The Fringe | Up to 29[4], see note 4 | Up to 32 | Up to 32 |
| Archaic ancestry | Up to 4[5], see note 5 | Up to 9[6], see note 6 | Up to 9[6], see note 6 |
| Polygenic scores | Up to 17 | Up to 17 | Up to 17 |
| Ancestry: maternal (mtDNA) | Up to 30[7], see note 7 | Up to 30[7], see note 7 | Up to 30[7], see note 7 |
| Ancestry: paternal (Y-DNA) | Up to 14[8], see note 8 | Up to 14 | Up to 14 |
| Ancestry composition | Up to 31 | Up to 31 | Up to 31 |
| APOE risk allele (e4) | Included | Included | Included |
| Lp(a) level | Included | Included | Included |
| Hemochromatosis (HFE) variants | Included[9], see note 9 | Included | Included |
| ABO blood type | Not available | Included | Included |
| APOL1 risk variants | Included[10], see note 10 | Included | Included |
| Celiac HLA types (DQ2.5 / DQ8) | Included[11], see note 11 | Included | Included |
| Deep Read: pharmacogenes from your reads, with CYP2D6 copy number, per-gene callability and HLA typing | Not available | Not available | Up to 29 |
| Recurrent deletion calls (from read depth) | Not available | Not available | Up to 2 |
[1]Note 1: ClinVar's known disease-causing variants (106,278 of them) in 1,363 recessive and X-linked genes and 52 dominant genes rated actionable; the listed variants are read, not the whole gene.
[2]Note 2: Array-dependent: the genes a chip can read in full, and only those whose defining SNPs your chip both typed and could resolve. DPYD needs a sequenced file.
[3]Note 3: 9 from a plain VCF; 25 from a gVCF, which also records the positions it examined. CYP2D6 copy number needs Deep Read.
[4]Note 4: Array-dependent: the most any chip can read, since a chip reads only single-letter markers it can place on the right strand. Which of them your chip reads depends on the array.
[5]Note 5: Array-dependent: the most any chip can read. The other markers are on no consumer array, or cannot be resolved from one, and which of these your chip carries depends on the array.
[6]Note 6: All of them from a gVCF, whose reference calls show each position was read. A plain VCF lists only where you differ from the reference, so it reads only the markers you carry.
[7]Note 7: Your maternal line is placed against the whole published mitochondrial tree, so the branch we name can be finer than this list. The count is the clades we have written about; a lineage outside it is still placed, and says so.
[8]Note 8: Needs your export's genome build, which 23andMe and AncestryDNA files give us. MyHeritage and FamilyTreeDNA exports don't state it, so for those we report the paternal line as unreadable rather than guess at it.
[9]Note 9: Array-dependent: a chip reads C282Y where it carries the position, and H63D only where it shows one copy. A sequenced file reads both.
[10]Note 10: Array-dependent: a chip reads G1 where it carries the position. G2 is a small deletion, which a sequenced file reads.
[11]Note 11: Array-dependent: both marker positions must be on your chip and readable from it. When they are not, the result reads as not called rather than as a negative.
A number is how many items we examine; means included; a dash () means that file type can't deliver it. Counts are the most a file of that type can give you, not a promise for every file.
These are curated, reviewed panels and a list of variants ClinVar already knows, not a comprehensive genome-wide screen, and the absence of a finding is not a clean result. A chip only reports the specific variants it typed, so its counts are an upper bound; full clinical findings, carrier status and the wider ClinVar screen need whole-genome data (WGS). Counts are version-stamped (2026-10-14) from the analysis registries. Polygenic scores are calibrated to broad ancestry groups and report your percentile against a reference population, not a personal probability of disease. Drop your file to see what yours covers.
Where Aimosti sits
| Service | Price | Updates | Input | Note |
|---|---|---|---|---|
| Genetic Genie | Free | Frozen | chip / VCF | Ad-supported, stale curation |
| Promethease | ~$25 | Frozen | chip / VCF | One-time SNPedia report, dense |
| SelfDecode | ~$99 / yr | Rented | chip (imputed) | Access stops when you stop paying |
| Aimosti | $39 once | Current & yours | whole-genome / chip | WGS-native (chip welcome), attributed, EU-sovereign |
Every other option hands you a one-time report frozen at the day you run it, or rents you ongoing access. With the subscription, Aimosti re-reads a whole-genome file after each new reference-data release it installs, and the report stays yours. On price it sits between the free report and the subscription. What the money buys is WGS-native depth, pinned curation you can re-check, and data that stays in the EU. And we never impute: where others fill in the genotypes a chip missed with a statistical guess, nothing in our pipeline invents a genotype. A no-call in your file reaches your report as a no-call.
One report. Yours to keep.
Read in Finland, attributed to the source, never sold.
See what your file covers