We don't have an opinion about your DNA. The literature does.
Every line in your report is a restatement of an authoritative, openly-licensed, version-pinned source. Here's exactly how that works.
The two-axis rule
We never collapse "how sure is the field?" and "how much does it matter?" into one score. That's how consumer genetics misleads. Every clinical finding shows evidence (ClinVar's own review stars) and effect (penetrance / effect size) side by side, so a big effect can never hide thin evidence, and vice-versa.
Pinned, attributed sources
Your report footer lists the exact version of every source it restates. Re-open it in a year and it renders identically: a frozen, attributable snapshot.
- ClinVar: variant–condition interpretations (we use ≥2★ by default)
- ACMG SF v3.3: the actionable-gene panel and its per-gene reporting rules
- CPIC: gene–drug prescribing guidance (pharmacogenomics)
- GWAS Catalog: the strict gate for wellness traits
- gnomAD: population frequency · PhyloTree: mtDNA lineages
The Frontier firewall
The Frontier module is where we relax the evidence bar to show you the interesting, biohacker-relevant variants: COMT and BDNF, alcohol-flush and caffeine metabolism, the natural short-sleeper genes, longevity curios, and the over-hyped classics we deliberately keep in to set the record straight (the “warrior gene”, the serotonin-transporter “depression gene”). We do not relax the sourcing bar (every entry still cites a real study), and we compensate with loud, mandatory caveats:
- Walled off from your clinical findings: a Frontier reading never feeds into or alters a clinical result
- A per-entry evidence-strength tier, from replicated down to contested and over-hyped
- A distinct visual register (never styled like a clinical finding) and a persistent "emerging research" banner
- A "what this isn't" line on every card