Biomarker knowledge graph · API · Early access

The cited reference layer for biomarkers.

What each biomarker measures, which laboratory tests cover it, what it relates to, and where that knowledge comes from. A graph your product can build on, with a source on every relationship.

Design-partner access first · One email, no marketing

5,284

biomarkers in the graph

3,648

laboratory tests, each linked to what it measures

1 : 1

relationships to sources. Nothing enters the graph without one.

What is in the graph

Organised around the biomarker, not the test menu.

Catalogs list tests. Papers discuss mechanisms. Guidelines publish ranges. Analyte Atlas connects them at the one node they share: the thing being measured.

Biomarkers: The centre of the graph: name, aliases, unit, LOINC codes.
Tests: Laboratory assays that measure a biomarker, with test code, specimen, preparation and the laboratory's own reference ranges.
Conditions: What a biomarker indicates, with ICD-10 codes where they exist.
Drugs and pathways: What acts on a biomarker, and through which mechanism.
Trials and research: Where the evidence comes from: registered trials and published papers.
Organizations and people: Laboratories, sponsors, investigators.

Provenance

Every relationship carries its source.

A relationship in Analyte Atlas is never a summary judgement. Each edge records where it came from: the catalog entry, the registry record or the paper, with the extraction that produced it and when. Quotes are rendered only when they match the source verbatim; otherwise you get the claim in our words and the link. If a fact cannot carry its source, it does not enter the graph.

Reference ranges are shown with the laboratory or guideline body that published them and the population they apply to. There is no single “normal range” in the graph, because there is none in the world.

A lookup, as the API returns it

GET https://api.analyteatlas.com/api/v1/biomarkers/creatinine

{
  "name": "Creatinine",
  "primary_unit": null,
  "alternative_names": ["cr","plasma creatinine","serum creatinine"],
  "measured_by": [
    { "test": "Zinc Random Urine",
      "lab": "Quest Diagnostics",
      "test_code": "16502",
      "specimen": "Urine",
      "loinc": [] },
    { "test": "5-Hydroxyindoleacetic Acid 5-HIAA Random Urine with Creatinine",
      "lab": "Quest Diagnostics",
      "test_code": "1648",
      "specimen": "Urine",
      "loinc": [] }
  ]
}

Live data from the current graph, trimmed for display

Endpoints, parameters and examples are on the API page.

Who it is for

Teams that would otherwise curate this by hand.

Consumer-lab and longevity platforms

A maintained biomarker-to-test-to-condition layer with citations, so your product can explain results and design panels without a team curating a spreadsheet.

Diagnostics companies and laboratories

Which assays measure which analytes, how panels compare, where a new assay fits, and the evidence behind a test's claims.

Pharma medical affairs and biomarker teams

Which biomarkers evidence which conditions and mechanisms, which trials used them, what the literature says, with sources ready for medical review.

Individuals are welcome too, on the public reference at Biomarker Lens, which renders the same graph as readable pages.

Access and pricing

Free to evaluate. Priced by use.

Evaluation

Read access with low limits, no redistribution. Enough to decide whether the graph fits your product.

API

Usage-based, by monthly request volume, with rate limits and support as the tiers.

Data licence

Annual, per organisation, for bulk access and use inside your own product.

The first design partners get a fixed annual fee, a defined scope and a direct line to the people building the graph. Write to hello@analyteatlas.com.

Build on a reference that shows its sources.

Request access and tell us what you are building. Design partners are chosen from this list.