GenomeNarrator: A Continuous, Auditable Personal Biology Engine
GenomeNarrator is not a one-time DNA report. Upload your 23andMe, AncestryDNA, MyHeritage, FTDNA, or VCF genome file once, including whole-genome and whole-exome sequencing (WGS/WES) VCFs up to 2 GB. The file is read in your browser, and only the genotype calls scoring needs are sent to our servers, which keep no copy. GenomeNarrator does not run the sequencing itself; it analyzes the VCF file you already have from any WGS/WES source (a clinical lab, Nebula Genomics, Sequencing.com, etc.), in addition to SNP-array exports from 23andMe and AncestryDNA. Your raw genome file never leaves your device, and no third party receives your genetic data. Every report combines two risk models: monogenic (ACMG/ClinVar single-gene findings like BRCA1/2) and polygenic risk scores (PRS) from GWAS + PGS Catalog for complex conditions like cardiovascular disease and type 2 diabetes, plus pharmacogenomics, ancestry, and a diet/exercise/sleep Lifestyle Blueprint. Then layer recurring bloodwork on top of that genome, any time, from a PDF, a FHIR R4 JSON export, or a scanned report read via on-device OCR, reconciled against your own genetic findings so an out-of-range marker comes with a gene-level reason, tracked on a chronological Timeline, with a plain note on what could not be checked. Everything that asks something of you is gathered into Your plan, one ranked checklist. Preview your results for free, or unlock a full report starting at $19.99 one-time or $14.99/month. No biology or genetics background is required; every finding is explained in plain English first, with the clinical detail available for anyone who wants to dig deeper.
How Genomic Analysis Works
Upload your genome file (the file never leaves your device) and the analysis checks your genotypes against more than 350,000 ClinVar variant records and more than 44,000 GWAS Catalog associations. No AI or large language model generates your results. Every finding is a deterministic match against published clinical data. Results include polygenic risk scores (PRS), pharmacogenomic drug interactions, ACMG v3.3 secondary findings across 84 genes, and ancestry breakdown. Learn more about our science and methodology.
Clinical Databases Behind Every Finding
Every variant in your report links to a peer-reviewed source:
- ClinVar: NIH's archive of clinical variant classifications from accredited labs worldwide. A single-gene finding needs at least two review stars (one for the 84 ACMG genes). Visit ClinVar at NCBI.
- GWAS Catalog: NHGRI-EBI catalog of published genome-wide association studies linking variants to traits and diseases. Visit the GWAS Catalog.
- CPIC: Clinical Pharmacogenomics Implementation Consortium guidelines for drug-gene interactions. Guidance for 133 gene-drug pairs across 37 genes and 4 HLA alleles, 54 of them at CPIC level A. Visit CPIC.
- gnomAD v4.1: Population allele frequencies from the Genome Aggregation Database, used to contextualise variant rarity. Visit gnomAD.
- PGS Catalog: EBI's catalog of validated polygenic score models, used to build ancestry-adjusted risk scores for complex conditions. Visit the PGS Catalog.
Every finding is cross-referenced against these independent sources rather than a single lookup; the citation is shown per-finding in your report so you can verify it yourself.
An Interactive Dashboard, Not Just a PDF
Your results don't just land as a static PDF; they open in a full interactive web dashboard. This is the primary way you use GenomeNarrator; the Compass PDF builder and Doctor/Clinical Summary PDF are optional exports generated from that same dashboard, not the primary experience.
- Your plan: one ranked checklist (Do first, Worth doing, Keep an eye on) of tests to confirm a DNA finding, out-of-range labs, affected medicines and everyday changes, each naming who acts on it.
- Nine sections in four groups: Home (an overview, then Your plan), Your DNA (Health risks, Medications, Lifestyle, Ancestry, Genes & Pathways), Your labs (Markers, Timeline) and PDF reports, each its own view rather than one long scroll.
- Interactive PGx Medication Checker: enter your own current medication list and cross-reference it live against every gene flagged in your pharmacogenomics results.
- Ancestry visualization: an ancestry composition breakdown across 8 gnomAD super-populations plus per-chromosome local-ancestry "chromosome painting," rendered as interactive charts, not static numbers.
- Data-freshness tracker: a live banner that quantifies exactly what's changed in ClinVar, the GWAS Catalog, and the PGS Catalog since your last analysis, with a link to re-analyze by re-uploading your file.
- Lifestyle tab: four pillars (Nutrigenomics, Exercise Physiology, Sleep and Chronobiology, Mind and Behavior) and a trait map of every curated trait with your result.
- Compass tab: build and customize your PDF report from inside the dashboard itself.
- 4 selectable visual themes (Native, Light, Dark, Forest) with live preview, so the dashboard matches how you like to read.
- Account panel: manage your subscription, security, and analysis history in the same interface.
Inside the Dashboard, Section by Section
The dashboard has nine sections in four groups: Home (an overview, then Your plan, one ranked checklist of what to do), Your DNA (Health risks, Medications, Lifestyle, Ancestry, Genes & Pathways), Your labs (Markers, Timeline) and PDF reports. A full tour with worked examples from the free sample report is at genomenarrator.com/llms-full.txt.
Home and Your plan. Three panels side by side (health risks, medicines, bloodwork), then One ranked checklist (Do first, Worth doing, Keep an eye on) across tests that would confirm a DNA finding, lab results out of range, medicines your genes affect and everyday changes. Each step says why in one sentence, who acts on it (your clinician, your prescriber or pharmacist, or you) and what to do next, with a tick box, a progress bar and a list to copy for your appointment. An Ask about your results box answers questions such as "Can I take codeine?" only from your own report: no AI-generated text, and nothing you type leaves the page.
Health risks. A genome summary (reference build, variants read, inferred sex, ClinVar release, average confidence); a risk radar placing every condition on six body-system axes; each condition's estimated absolute risk above or below the population average; hereditary cards with the gene, rsID, ClinVar classification and review stars, a next step, and what the gene's inheritance means for one copy; a very rare result read from a DNA chip labelled an unconfirmed chip call and never ranked urgent; and every condition grouped by body system.
Medications. How many common medications your genes affect (avoid, dose change suggested, monitor, not flagged), which genes drive it and how many of each gene's markers your file carried, a Medicine Cabinet grouped by therapeutic area, the lab tests that matter for those medicines, a checker for the drugs you take, and one card per pharmacogene. Signed-in members can save their medication list and are told when its guidance changes.
Lifestyle. A 6 AM to midnight day timeline of the results worth acting on, then four pillars (Nutrigenomics, Exercise Physiology, Sleep & Chronobiology, Mind & Behavior), each with the genes it reads, a whole-genome polygenic score with its percentile, PGS Catalog model ID and how many of the model's variants your file carried, and your result for each of 121 curated traits.
Ancestry. The inferred population and subgroup the report is calibrated to, composition with 95% intervals, the founder-variant panels screened, the mtDNA haplogroup (and Y-DNA for males) with its ancestral path, and chromosome painting across all 22 autosomes.
Genes & Pathways. For each gene behind a finding: gnomAD gene constraint (pLI, LOEUF and Z-scores), its top GTEx tissue, Reactome pathways, a network of genes that share a pathway, and a tissue-expression map.
Markers. Bloodwork from a lab PDF (Quest, LabCorp, Function Health and similar), a scanned report, a FHIR file or typed values, 80 tests in 11 panels, read against your genome. Your Results comes first: each out-of-range value on a Low/Normal/High gauge with your previous test, the normal range, an expected-for-your-genotype band where a published study gives the gene's effect, and one line naming the gene. Then Tests to consider, ordered by what a result would change; your lab history; and Patterns Across Your Results, with six numbers worked out from results you already have (HOMA-IR, triglyceride to HDL, non-HDL cholesterol, ApoB against LDL, transferrin saturation, metabolic syndrome criteria). Clinician Summary and FHIR exports.
Timeline. Every lab draw on one time axis in its own units with its normal band; a change drawn solid only when it exceeds that marker's reference change value from peer-reviewed biological variation, dashed when it cannot be tested yet; a Next draw column; and your genes drawn as arrows to the markers they predict.
PDF reports (Compass). A PDF report with a Download Full Report button first and chapters for Chronic Disease, ACMG, Diet, Exercise, Sleep, Neurobehavioral, Supplements and Traits, each with a preview of its pages.
On a phone the same dashboard has a bottom tab bar and a strip of each tab's sections under the header.
What's In Your Report: Every Claim, Sourced
These are GenomeNarrator's own specifications, not a comparison against other tools. Each row below cites the primary source used to verify it.
| Capability |
What GenomeNarrator does |
Verify it |
| Plain-English Translation | Every finding is explained in everyday language first (what it means and why it matters), with the underlying ClinVar/CPIC/ACMG clinical data shown underneath for anyone who wants to verify it, never required reading. Not a raw database lookup. | Shown per-finding in your report |
| Data Sources | ClinVar, the GWAS Catalog, the PGS Catalog, gnomAD v4.1, CPIC, AlphaMissense and the ACMG SF list, cross-referenced per finding rather than a single-source lookup | Our methodology |
| Pharmacogenomics (PGx) | 37 genes and 4 HLA alleles · 133 gene-drug pairs, 55 at CPIC level A, plus an interactive Medication Checker that matches your own current drug list against every flagged gene | CPIC guidelines |
| Whole Genome / Exome Sequencing (WGS/WES) | VCF files up to 2 GB accepted and analyzed directly, in GRCh37 or GRCh38, not limited to SNP-array data | Upload a WGS/WES VCF |
| ACMG Secondary Findings | All 84 genes on the official ACMG SF v3.3 panel, screened: actionable findings paired with ClinGen-sourced next steps (surveillance, preventive options, medication guidance), including a dedicated surgical-anesthesia alert for RYR1/CACNA1S malignant hyperthermia carriers | Lee et al. 2025, Genetics in Medicine |
| Polygenic Risk Scores (PRS) | 139 conditions scored with 97 published PGS Catalog models: cardiovascular, type 2 diabetes, and other complex/metabolic conditions | PGS Catalog |
| Bayesian Absolute Risk | Ancestry-matched priors from gnomAD v4.1 (807,000+ individuals) | Our methodology |
| Ancestry Composition | Auto-inferred across 8 gnomAD super-populations from ancestry-informative markers, with a confidence score, 95% CI, and sub-population refinement | Our methodology |
| Local Ancestry (Chromosome Painting) | Per-chromosome ancestry-segment painting via HMM & Viterbi decoding, validated to 89–94% window accuracy for high-divergence population pairs, 70–84% for closely related ones | Our methodology |
| Maternal / Paternal Haplogroups | mtDNA (PhyloTree Build 17) and Y-DNA (32 curated ISOGG-defining SNPs) haplogroup calls with confidence scoring, ancestral lineage, and migration path | Our methodology |
| Carrier / Compound-Het | Zygosity-aware carrier flags + phase-checked compound-het calls | Our methodology |
| Population / Founder Carrier Panels | 8 ancestry-specific founder-mutation panels (Ashkenazi Jewish, Finnish, African, South Asian, Middle Eastern/North African, East/Southeast Asian, European, Americas/Latino) covering 56 curated founder variants, auto-screened once your inferred ancestry for that population reaches 5%+ | Our methodology |
| Traits & Lifestyle Blueprint | 121 curated traits; 7 carry evidence-graded (A/B/C) diet, exercise, sleep, and stimulant-metabolism recommendations, 21 in all, one set for each possible result | Trait catalog |
| Genome File Stays on Your Device | Your raw genome file is read in your browser and never uploaded. Only the genotype calls scoring needs are sent to our servers, and none are stored | How this works |
| Evidence Audit Trail | Every finding cites its source record: ClinVar accession, GWAS study, or PGS ID | Shown per-finding in your report |
| Finding Confidence | Every finding carries an Evidence Grade (0–100) and a ClinVar review tier: Expert Panel, Multiple Submitters, or Single Submitter | How grading works |
| Database Currency | ClinVar + GWAS Catalog rebuilt monthly; gnomAD, PGS Catalog, and CPIC refreshed each major release; your dashboard quantifies exactly what's changed since your last analysis, not just a generic reanalyze prompt | Our data sources |
| Interactive Dashboard | Results open in a full interactive web dashboard, not just a static PDF: nine sections in four groups: Home (an overview, then Your plan, one ranked checklist of what to do), Your DNA (Health risks, Medications, Lifestyle, Ancestry, Genes & Pathways), Your labs (Markers, Timeline) and PDF reports, with the interactive PGx Medication Checker, ancestry composition plus per-chromosome local-ancestry chromosome painting, a data-freshness tracker that shows what changed since your last analysis, a Lifestyle tab, a built-in Compass PDF builder, and 4 selectable visual themes with live preview | See it live: free demo |
| Genes & Pathways | Every gene behind a finding gets its own drill-down: shared-pathway convergence across genes, GTEx tissue-expression body map, gnomAD gene-constraint (pLI/LOEUF) scoring, and expression-QTL data, a sortable cross-gene summary table plus an expandable per-gene card grid | See it in the dashboard |
| Blood Lab Panels | 80 blood tests across 11 panels: blood count with white-cell differential, lipids including ApoB and Lp(a), glucose and insulin, liver, kidney and electrolytes, thyroid, iron, vitamins, inflammation, muscle enzymes and hormones, read under the names Quest, LabCorp and Function Health print, with trend tracking and every value shown beside its flag and its normal range | See it in the dashboard |
| Lab Ingestion Formats | Bloodwork parsed from a standard PDF, a FHIR R4 JSON export (Bundle/Observation resources), or a scanned/photographed report read via on-device OCR, all parsed client-side and never uploaded raw, or values typed in by hand with each one flagged against its reference range as it is entered | See it in the dashboard |
| Genetics x Labs Crossover | 85 cited links from 64 genes to 45 of the 80 tests, each carrying its evidence grade, source and last review date; 47 also state how far a genotype shifts the average value, drawn as an expected-for-your-genotype band. A result is read as matching the genetic prediction, opposite to it, or in range, and a gene that was checked and does not carry the relevant variant is reported as not the explanation, so a flagged value points elsewhere instead of being pinned on DNA | See it in the dashboard |
| Biological Map | For every gene-to-lab link: what the gene does and the organ it acts in, with its GTEx v10 expression there and its Reactome pathway where the reference data has them. The organ is chosen for the biology of that lab value, not the tissue where the gene happens to be most expressed, and every expression figure and pathway is checked against the reference data by automated tests | See it in the dashboard |
| Tests to consider | Which blood tests are worth considering next, ranked by what a result would actually change for the person rather than by how alarming a gene sounds. Tests that only matter if a specific medication is started are listed separately and never pushed as a reason to test now | See it in the dashboard |
| What Could Not Be Checked | Under the Your Results table, a plain sentence says what could not be checked: genes checked and clear, and results with no blood marker linked yet. Tests to consider ranks the next tests by what a result would change, not a generic reminder to get more labs done | See it in the dashboard |
| Your plan | One ranked checklist across tests to confirm a DNA finding, out-of-range labs, medicines your genes affect and everyday changes, each step naming who acts on it, with tick boxes and a list to copy for your appointment. An Ask about your results box answers from your own report only, with no AI-generated text | See it in the dashboard |
| Population Reference Bands | NHANES 2017-2018 population reference-interval bands rendered behind a person's own lab trend line, so an individual result can be read against a real population distribution, not just a single lab's flag threshold | See it in the dashboard |
| Clinician Summary (Genetics × Labs) | A separate 1-page export distinct from the Doctor/Clinical Summary PDF, bucketing the genetics×labs cross-reference into Concordances, Discordances, and Actionable Unknowns for a physician visit | Full pricing |
| Compass Report Builder | Customizable PDF report engine, available from the Essentials tier, covering every body system and finding in your analysis, not a fixed static export | Full pricing |
| Doctor / Clinical Summary PDF | Separate multi-page physician summary covering ACMG SF findings, PGx phenotypes, and pathogenic variants, built for sharing with a doctor, not a raw data dump | Full pricing |
| Try Before You Buy | A bundled demo genome unlocks the full Complete-tier report instantly in your browser, no signup, no payment, no DNA file required | Try the demo |
| Raw Data Export (CSV) | Every condition and trait match exportable as spreadsheet rows: risk category, confidence, variant count, matched RSIDs | Full pricing |
| Price | $19.99 one-time (30-day access) or subscriptions from $14.99/mo that keep the full report open | Full pricing |
Pharmacogenomics: How Your Genes Affect Medications
The pharmacogenomics section identifies how your CYP2C19, CYP2D6, VKORC1, DPYD, and SLCO1B1 variants affect your response to common medications including warfarin, clopidogrel, SSRIs, statins, and chemotherapy agents. All findings follow CPIC Level A and B guidelines. Read our pharmacogenomics guide to understand how star alleles and phenotype predictions work.
Disease Risk Across 1,000+ Conditions
Your report covers cardiovascular disease, cancer predisposition (including BRCA1/2 and Lynch syndrome via ACMG secondary findings), neurological conditions such as Alzheimer's disease (APOE), metabolic disorders such as type 2 diabetes, autoimmune conditions, and rare Mendelian diseases. Complex, multi-gene conditions are scored with polygenic risk scores (PRS) computed from validated GWAS associations, not just single-gene lookups. Learn how genome-wide association studies translate into risk estimates.
Privacy by Architecture
GenomeNarrator was built from day one so that your raw genome file itself never touches a server. It is parsed and read directly in your browser. There is no storage of your raw DNA, and no third-party analytics that touches your genetic data. See our privacy policy for the full details.
Pricing
GenomeNarrator offers a free preview (2 body systems, 3 trait insights, and an ancestry summary) so you can see the pipeline before purchasing. Full access to all 1,000+ conditions, pharmacogenomics, and ACMG secondary findings is available as a $19.99 one-time 30-Day Pass (no subscription), or as an ongoing Essentials subscription from $14.99/month ($119.99/year). Plus ($24.99/month) adds the full Traits tab and the combined Compass PDF with every chapter, and Complete ($34.99/month) adds the Lifestyle Blueprint and a Doctor/Clinical Summary PDF. See full pricing for all five tiers.
Frequently Asked Questions
What genome file formats does GenomeNarrator support?
We support raw data files from 23andMe (.txt), AncestryDNA (.txt), MyHeritage, FTDNA, and standard VCF files in GRCh37 or GRCh38 format, including whole-genome and whole-exome sequencing (WGS/WES) VCFs. Files up to 2 GB are supported, and large WGS VCFs are streamed and parsed client-side rather than loaded into memory all at once. You do not need a premium subscription with your provider; just download the raw data file.
Do I need a background in biology to understand my results?
No. GenomeNarrator is built for people without a genetics background. Every finding starts with a plain-English explanation of what it means and why it matters, before any clinical terminology. ClinVar accessions, CPIC guideline levels, and ACMG classifications are shown for anyone who wants to verify a finding independently, but they are never required reading to understand your own report.
Is my genetic data private and secure?
Yes. Your raw genome file is parsed in your browser and never leaves your device; it is never uploaded. No third-party access to your raw DNA. We never sell or share your genetic data.
How accurate is the genomic analysis?
A single-gene finding needs a Pathogenic or Likely Pathogenic ClinVar record with at least two review stars (one for ACMG genes), and medication guidance follows CPIC guidelines. Polygenic risk scores use validated GWAS associations from the NHGRI-EBI Catalog. SNP arrays only detect common, pre-selected positions; rare pathogenic variants and structural changes need whole genome sequencing (WGS) coverage to be assessed at all. If you already have a WGS or WES VCF file, GenomeNarrator analyzes it directly and gets that deeper coverage; if you only have 23andMe/AncestryDNA data, a clinical WGS panel is the way to close that gap. Drug-response gene calls agree with PharmCAT, PharmGKB's reference caller, on 99.4% of 46,341 gene calls in 3,202 people from the 1000 Genomes Project.
Do I need a 23andMe or AncestryDNA account to use GenomeNarrator?
No. You only need the raw data file, which you can download directly from your testing provider's website under account settings or data downloads. The analysis works with any raw genome file in the standard tab-separated SNP format.
Does GenomeNarrator calculate polygenic risk scores (PRS)?
Yes. Alongside monogenic ACMG/ClinVar findings for single-gene conditions like BRCA1/2, 139 complex conditions, cardiovascular disease and type 2 diabetes among them, are scored with 97 published PGS Catalog models, and each score says which populations it was built and tested in. Where a published baseline prevalence exists (270+ conditions), the result is also given as an absolute risk, adjusted against gnomAD allele frequencies for your inferred ancestry.
Does GenomeNarrator give diet, exercise, or lifestyle recommendations?
Yes. The Lifestyle Blueprint (Complete plan) turns 7 curated traits into evidence-graded (A/B/C) recommendations across five pillars: Nutrigenomics (diet & nutrition), Exercise Physiology, Sleep & Chronobiology, Stimulant Metabolism, and Taste & Temperament.
How much does a genomic analysis cost?
You can preview your results for free. Full reports start at $19.99 for one-time 30-day access, or from $14.99/month as a subscription; Plus adds the full Traits tab and one combined PDF of every chapter, and Complete adds the Lifestyle Blueprint and a doctor-ready summary PDF. Visit our pricing page for all five tiers.
Is GenomeNarrator just a PDF report, or is there an interactive dashboard?
Both, and the dashboard comes first. After you upload a genome file, or try the free demo, your results open in a full interactive web dashboard of nine sections in four groups: Home (an overview, then Your plan, one ranked checklist of what to do), Your DNA (Health risks, Medications, Lifestyle, Ancestry, Genes & Pathways), Your labs (Markers, Timeline) and PDF reports, with an interactive PGx Medication Checker, ancestry composition and per-chromosome local-ancestry chromosome painting, a data-freshness tracker, a Lifestyle tab for diet, exercise and sleep, and 4 selectable visual themes. The Compass PDF builder and a separate Doctor/Clinical Summary PDF are optional exports generated from that same dashboard, not the primary experience.
What is the difference between a SNP array and whole genome sequencing? Does GenomeNarrator support WGS?
SNP arrays (used by 23andMe and AncestryDNA) test around 600,000 to 700,000 pre-selected genetic positions. Whole genome sequencing (WGS) reads every base pair in your genome. Yes, GenomeNarrator directly supports WGS: if you already have a WGS or WES VCF file (from a clinical lab, Nebula Genomics, Sequencing.com, or any other sequencing provider), you can upload it to GenomeNarrator and it is analyzed the same way as a 23andMe or AncestryDNA file, files up to 2 GB. The file is read in your browser and never uploaded. GenomeNarrator does not perform the sequencing itself (it is not a saliva-kit lab); it analyzes the VCF a WGS/WES provider gives you. For most common disease risk and pharmacogenomics, SNP array data is sufficient; for rare disease diagnosis, BRCA confirmation, or comprehensive structural variant analysis, a WGS VCF gives deeper coverage than an array file. Read our guide on how genomes work for more.
Does GenomeNarrator use artificial intelligence to generate results?
No. Every finding is a deterministic lookup against ClinVar, CPIC, and the GWAS Catalog. No large language model or AI generates text in your report. The pipeline is open to review on our science page.
Is GenomeNarrator a one-time DNA report or an ongoing service?
GenomeNarrator is a continuous, auditable personal biology engine, not a one-time report. ClinVar and the GWAS Catalog are rebuilt monthly and gnomAD/PGS Catalog/CPIC are refreshed each major release. Your genome file is not stored on our servers, so a report is re-scored when you re-upload the file; your dashboard shows what has changed since your last analysis, and if ClinVar reclassifies a variant your report showed, your saved report is flagged and you get an email. You can also layer in new bloodwork at any time; each lab result is reconciled against your existing genetic findings rather than replacing the report.
What happens when I upload a new blood test?
The lab result, from a PDF, a FHIR R4 JSON export, or a scanned report read via on-device OCR, is parsed entirely client-side, cross-referenced against your own genetic findings for a gene-level explanation of any out-of-range value, and added to your Timeline and Your plan.
How does the Markers page say what it could not check?
Under the Your Results table, a plain sentence lists what could not be checked: the genes that were checked and are clear, and the results that have no blood marker linked yet. Tests to consider separately ranks the next tests worth considering by what a result would change, rather than giving a generic "get more labs done" prompt.
Ready to understand your genome? Start your analysis. Upload takes under a minute; preview for free, or unlock the full report for $19.99 one-time or from $14.99/month.