Use case · Biomarker discovery

Find the right biomarkers, with the evidence behind every candidate

Causaly brings together the literature, Bio Graph and genetic and cancer genomics data, so your team can surface biomarker candidates, compare them on the same criteria and check every claim at its source.

Why use Causaly

A shortlist of biomarkers you can defend, not a pile of papers

Causaly surfaces candidates with the evidence behind each one, compares them on the same criteria and is clear about what is established, so fewer surprises reach the lab.

Candidates, not just papers

Causaly extracts the molecules linked to a disease, mechanism or drug response, with the supporting papers and graph relationships for each.

The same criteria for every candidate

Disease linkage, biomarker class, tissue expression and the strength of preclinical and clinical evidence are set out the same way each time.

Clear about what is established

Reports show the type and strength of evidence behind each candidate, and where clinical evidence is missing, so nothing looks more established than it is.

What you get

Candidates worth validating, and the case for each

Each candidate comes with the evidence for and against it, drawn from Bio Graph, the literature and connected genetic and cancer genomics databases, helping you decide which candidates deserve investment.

648M+

scientific relationships across Bio Graph and Pipeline Graph

40M+

biomedical publications and preprints

550+

public cancer genomics studies available through cBioPortal

225K

GWAS variants with gene context

Content as of October 2026. New literature reaches Bio Graph within about four weeks of publication.

How it works

How Causaly helps you find and compare biomarkers

Ask a question about a disease, target or drug in Agentic Research, or run a library workflow such as Biomarker Assay Selection.

  • Input

    A disease, target or drug, and what the biomarker is for

    For example diagnosis, patient stratification, monitoring or response to treatment.

  • Evidence

    Literature, Bio Graph and genomics data

    Causaly searches Bio Graph and the literature, plus connected sources such as gnomAD, ClinVar, Human Protein Atlas, GEO and cBioPortal, and links each finding to the record it came from.

  • Assessment

    Candidates compared on the same criteria

    Each candidate is set out with its disease link, expression, genetic support and the strength of preclinical and clinical evidence.

  • Output

    A cited shortlist and a measurement plan

    Biomarker Assay Selection evaluates the assay evidence and recommends a fit-for-purpose measurement strategy. Reports export to Word, PDF or PowerPoint.

Key capabilities

What biomarker teams can do with Causaly

Teams use Causaly to move from a long list of possible biomarkers to a short list they can defend, with the evidence for each decision in one place.

Surface candidates for a disease or target

Find genes, proteins and other molecules linked to a disease, a mechanism or a drug response, with the supporting papers and graph relationships for each.

Check genetic and cancer genomics support

Add population allele frequencies and loss-of-function constraint from gnomAD, and mutation, copy-number, expression and clinical data from cBioPortal studies.

Compare candidates on the same criteria

Weigh disease linkage, biomarker class, tissue expression and the strength of preclinical and clinical evidence, so every candidate is assessed the same way.

Choose how to measure them

The Biomarker Assay Selection workflow evaluates the evidence for each assay option and recommends a fit-for-purpose measurement strategy before you commit lab time.

Find response, resistance and safety signals

Look for biomarkers of treatment response, resistance and toxicity across drugs and indications, with each finding traced to its source.

Analyze your own data

With Causaly Work, in public preview, give a task such as a differential expression analysis and get back a finished deliverable, with the code, parameters and evidence alongside.

Optional: connect internal data through the Enterprise Data Fabric or your own MCP connectors, so your team’s findings are read alongside the public evidence.

“I use Causaly to identify suitable biomarkers and get an idea of feasibility for use in first-in-human dose trials and Phase 1 trials. Manual research in PubMed and Google takes significantly longer and the risk of missing out on something important is much greater.”
Senior Scientist · Top 20 pharma

77%

of scientists say Causaly improves their decision-making

40%

average time savings observed

$26–42M

potential cost savings for a typical organization of 5,000+ scientists

Source: Causaly customer survey. Cost savings reflect efficiency gains and faster deprioritization of non-viable targets.

Biomarker discovery: common questions

Which data sources does Causaly use for biomarker research?

Causaly draws on Bio Graph and Pipeline Graph, with 648M+ scientific relationships, alongside 40M+ publications and preprints. Agentic Research can also query connected sources such as gnomAD, Human Protein Atlas and cBioPortal, among others.

Does Causaly validate biomarkers?

Causaly assesses the published and structured evidence for a biomarker and shows how strong it is. Analytical and clinical validation still happen in the lab and in studies. Causaly helps your team decide which candidates are worth that investment.

Can I see the evidence behind each candidate?

Yes. Every claim in an answer or report links to its source, so you can open the paper or record and check it yourself before you make a decision.

Can Causaly help choose an assay for a biomarker?

Yes. The Biomarker Assay Selection workflow evaluates the evidence for each assay option and recommends the most credible fit-for-purpose measurement strategy.

Can I export biomarker reports?

Yes. Answers and reports can be exported to Word, PDF, PowerPoint or Excel and shared with your team.

Bring a biomarker question your team is working on

We’ll run it with you and walk through the candidates, the evidence for each and how to measure them.