It's time to build, on the genome.

We handle everything from sequencing to a continuously interpreted genome, so you can build personalised applications for every user.

Personalisation is still built on averages.

Most products personalise from demographics, surveys, behaviour, wearables, or laboratory results. These signals are useful, but they mostly describe what has already happened.

The genome helps explain what population averages miss: why two people can eat the same food, take the same medication, follow the same training plan, or share the same environment and respond differently. It gives every other signal individual context, so products can move from grouping similar users to understanding the person in front of them.

Different by default

People with the same goal and the same inputs can have meaningfully different biological responses.

Lifelong context

A genome is read once and can add context across many features and stages of a user's life.

More useful over time

Hosted genomes are automatically reinterpreted as the underlying evidence and knowledge base change.

We abstract the complexity of the building on the genome.

Collection, laboratories, large genomic files, variant processing, scientific interpretation, evidence updates, and safe user-facing outputs are normally separate systems. The Genome API turns them into one product surface.

Sequencing and import

Order a collection kit or use an existing gVCF. The API covers shipping, laboratory processing, conversion, and secure delivery.

3x WGS · 30x WGS · GSA · existing gVCF

Structured delivery

Sequencing data is returned as a structured .genome bundle, with conventional gVCF and FASTQ deliverables available where applicable.

.genome · gVCF · FASTQ · hosted genome

Hosted genomes

Hosted genomes are automatically reinterpreted as the underlying research and knowledge base changes. Each interpretation remains versioned.

automatic reinterpretation · version history · provenance

Interpretation outputs

Applications can request structured panels or bounded plain-language answers with evidence, citations, limitations, and safety metadata.

14 panels · Q&A · webhooks · scoped access

2 SDKs

TypeScript and Python

14 panels

Structured interpretation outputs

1 API

Sequencing through interpretation

Test with synthetic genomic data.

Staging includes fixed synthetic genomes for integration tests. Production uses the same API surface with separate credentials, data controls, and billing.

Five-minute sandboxTypeScript

import { GenomeComputer } from "@genomecomputer/sdk";

const gc = new GenomeComputer({
  apiKey: process.env.GC_API_KEY!,
  environment: "staging"
});

const panel = await gc.genomes.buildPanel(
  "gen_test_nutrition_001",
  "nutrition_diet_response"
);

console.log(panel.summary);

One downstream model

A genome created through sequencing and one created from an uploaded gVCF use the same hosted genome, interpretation, panel, and Q&A resources.

New sample3x · 30x · GSA

Existing gVCFsecure upload

Genome runtimeversioned · queryable

Panels + answersevidence · citations

Example applications

Panels and bounded Q&A can support common application categories. Lower-level interpretation artifacts and provenance remain available for expert and audit views.

Consumer health

Structured panels are available for nutrition, sleep, fitness, recovery, hormone, skin, and healthy-aging applications.

Medication-related genetics

Pharmacogenomic outputs can support clinician-facing workflows with provenance and explicit limits around diagnosis and dosing.

Agents and software tools

Agents can query a deterministic genome resource rather than extract genetics from reports or loosely encoded variant files.

Files, hosting, and provenance

Results can include a portable, provenance-pinned .genome bundle alongside conventional deliverables. Hosted genomes add automatic reinterpretation and persistent API access.

Read the API reference

Create a production order

const order = await gc.orders.create(
  {
    type: "wgs_30x",
    retention: "hosted_runtime",
    deliverables: ["genome_bundle", "gvcf", "fastq"],
    customer: {
      name: "Avery Chen",
      email: "avery@example.com",
      phone: "+1 415 555 0100"
    },
    shipping_address: {
      line1: "535 Mission St",
      city: "San Francisco",
      state: "CA",
      postal_code: "94105",
      country: "US"
    }
  },
  { idempotencyKey: "order-user-123" }
);

Provenance

Versioned interpretations, evidence levels, citations, limitations, and source provenance travel with the result.

Safety boundaries

Panels and answers are bounded: no diagnosis, dosing, or prescriptions. Clinical decisions remain with clinicians.

Corporate structure

Genome Computer operates under Genetic Superintelligence Company, a Public Benefit Corporation. Identifiable individual genetic data is not sold or licensed.

Request API access

Provide your intended use case and expected volume. We will follow up about staging access, production requirements, and pricing.

Prefer email? api@genome.computer