Methodology & accuracy
How PeptidesDNA matches peptides to your DNA
PeptidesDNA is built specifically to interpret your DNA for peptide response. It scores 39 peptides against a 144-marker panel and 8 pharmacogenes — the peptide-specific depth that general ancestry and wellness DNA reports, which treat peptides as a footnote, do not provide. This is exactly how the scoring works, and why a dedicated platform goes deeper on this one question.
The short answer
PeptidesDNA analyzes a 144-marker peptide-relevant panel and an 8-gene pharmacogene panel. Each peptide gets a fit: how much of that compound’s own genetic case your file matches, shown next to how many of its markers we could actually read, so a result from two markers never looks like a result from fifteen. Markers your file could not read are excluded rather than guessed. The pharmacogene calls follow CPIC and PharmGKB star-allele definitions. The peptide corpus is built from published genetic-association literature and carries an evidence grade we assign and publish, because no pharmacogenomic guideline covers peptides yet. Scoring is deterministic: the same file and the same corpus always give the same numbers.
144
Marker panel analyzed
8
Pharmacogenes called
39
Peptides scored
3
Evidence tiers
How the score is built
Every peptide is linked to a set of genetic markers, each carrying a weight that reflects how strong the published evidence behind it is. Your score is built from the markers you actually carry.
Count what you carry
For each marker linked to a peptide, we check your genotype and count whether you carry one copy of the relevant allele or two. Each hit adds its weight to that peptide's raw total.
Score against the compound's own case
Each peptide is scored against the most its own markers could score, so the number is how much of that compound's genetic case you match. It is a statement about the fit between you and that compound, not a rating of the compound and not a comparison against other people.
Show per-peptide fit separately
Each card also shows how much of that specific peptide's own genetic case you hit, which is a different question from where it ranks. A peptide with few known markers can be a strong personal fit and still rank low.
Exclude what we could not read
Markers your file did not cover are left out of both the total and the maximum, so a gap never quietly counts as a miss. Uncovered markers are listed for you by name.
Evidence tier is reported alongside every result rather than folded into the number, so a high score never hides thin science.
The scoring is deterministic — the same DNA always yields the same scores. Narratives are generated from those fixed numbers, so there is no hallucination risk on the values themselves.
Why a dedicated platform goes deeper than a general DNA test
A broad ancestry or wellness report may genotype millions of markers, but only a handful are relevant to peptides, and it never scores them against a peptide's mechanism or adjusts dose by your metabolizer status. PeptidesDNA does only this — so it goes deeper where it counts.
| Capability | PeptidesDNA | General DNA report | Manual research |
|---|---|---|---|
| Built specifically for peptides | Yes | No — a minor feature | No |
| Peptide-relevant markers analyzed | 144-marker panel | Few / incidental | — |
| Pharmacogene panel (8 genes) | Yes | Sometimes, generic | No |
| Scored against each peptide's mechanism | Yes | No | Manual |
| Metabolizer context for dosing discussions | Yes | No | No |
| Evidence tier on every result | Yes | No | No |
| Peptides scored | 39 | 0–few | DIY |
| Continuously updated library | Yes | Rarely | — |
Comparison of PeptidesDNA against general consumer DNA reports and unaided manual research, on peptide-specific capabilities.
Frequently asked questions
How is PeptidesDNA different from a general DNA test for peptides?
It's built specifically for peptides. PeptidesDNA genotypes a 144-marker peptide-relevant panel and 8 pharmacogenes, then scores each of 39 peptides against the mechanism it works through. General ancestry or wellness DNA reports treat peptides as one small feature and don't carry this peptide-specific depth.
How many genetic markers does PeptidesDNA analyze?
A 144-marker panel across 15 biological categories — receptor variants and pathway, inflammation, metabolic, and transporter markers — plus an 8-gene pharmacogene panel: CYP2C19, CYP2C9, CYP2B6, CYP3A5, VKORC1, TPMT, SLCO1B1, and CYP4F2. Consumer arrays cannot reliably resolve CYP2D6, so we don't report it.
How is each peptide scored?
Each peptide is linked to a set of genetic markers, each weighted by how strong the published evidence behind it is. We add up the markers you actually carry, counting one copy or two, and divide by the most that peptide's own markers could score. The number on a card is therefore how much of that compound's genetic case your file matches, not a ranking of you against other people. Markers your file could not read are excluded rather than guessed. Where two markers sit close enough together to be inherited as a single signal, we count them once, using linkage patterns measured in European-ancestry reference data: that is the best characterised panel available, and it is a limitation worth knowing if your ancestry is different. Anything we could measure at fewer than three markers sits at the bottom of the list however well it scored, because a handful of markers cannot tell a real match from luck. Every result carries an evidence tier, Established, Emerging, Preclinical or Withdrawn, so confidence is never hidden, and the scoring is deterministic, so the numbers are reproducible rather than generated by a language model.
What data sources does PeptidesDNA use?
Two different corpora, and they rest on different footings. The 8-gene pharmacogene panel calls star alleles against CPIC and PharmGKB definitions, which is the established standard for those genes. The peptide-matching corpus does not have that luxury: CPIC publishes 29 dosing guidelines and none of them covers a peptide, so that corpus is built from peer-reviewed genetic-association literature and graded by us. We would rather say that than imply a guideline exists where none does.
Can I use my existing 23andMe or AncestryDNA data?
Yes. PeptidesDNA is upload-first: upload your raw file from 23andMe, AncestryDNA, MyHeritage, or any major provider and get your report within 24 hours for $99. Existing consumer files typically cover 85–95% of the 144-marker panel; any uncovered markers are flagged, not guessed.
Are there other DNA-to-peptide services?
Yes, a few exist, and some general DNA-report services include a small 'response to compounds' section. What sets PeptidesDNA apart is focus: the entire engine, marker panel, and research library are built for one question — which peptides relate to your DNA — with a transparent evidence tier on every result.
Data sources
Pharmacogenes. Star alleles are called against CPIC and PharmGKB definitions, cross-checked in dbSNP. That is the established standard for these genes, and it is why we report them as Indeterminate rather than guessing when a defining marker is missing from your file.
Peptides. There is no equivalent standard, and we would rather say so than imply one. CPIC publishes 29 dosing guidelines and not one covers a peptide. So this corpus is built from peer-reviewed GWAS and peptide-pharmacology literature, cross-referenced against the GWAS Catalog, Open Targets and ClinPGx, and every compound carries an evidence grade we assign, publish and can be held to. Where a marker is read and the research does not support a call, the report says exactly that instead of manufacturing a finding.
PeptidesDNA provides genetic interpretation for educational purposes and is not medical advice.
See your peptides ranked by your DNA
Upload your existing 23andMe, AncestryDNA, or MyHeritage data and get your personalized report within 24 hours.