DYSPLAI Python SDK

For Research Use Only. Not for use in diagnostic procedures.

DYSPLAI is a molecular oncology research platform spanning sequence analysis, reference-atlas positioning, cohort discovery, molecular evidence graphs, counterfactual pathway exploration, clinical metadata harmonisation and reproducible reporting.

This is the typed Python client for it: Pydantic v2 response models, a blocking wait() helper for long-running jobs, and thirteen service surfaces. The SDK is a thin client — every model, atlas and pipeline stays on DYSPLAI infrastructure.

Install

pip install dysplai
# with notebook dependencies:
pip install "dysplai[notebooks]"

Quick start

from dysplai import DysplaiClient

with DysplaiClient(api_key="your-key") as client:
    # Submit a molecular analysis
    job = client.analysis.submit(
        input_uri="gs://your-bucket/sample.csv",
        sample_id="SAMPLE_001",
        cancer_type_hint="lung_adeno",
    )

    # Wait for analysis to complete (~5–15 min)
    result = client.analysis.wait(job.analysis_id, poll_interval=10)

    print(result.cancer_type_probabilities)
    print(result.biological_neighbourhood_description)
    print(result.ruo_disclaimer)

API surfaces

The molecular evidence graph (client.graph) grounds a prediction in the evidence behind it — supporting and contradicting — rather than returning a score alone. Graph reads take the analysis id and are tenant-private.

Worked examples

Task-oriented guides — cohort discovery, evidence graphs, counterfactual exploration, imaging, assay design and reporting — are in the developer documentation at https://www.dysplasiadx.com/developers, alongside the Colab notebooks listed below.

Error handling

from dysplai import AuthError, QuotaError, NotFoundError, ValidationError

try:
    result = client.analysis.result("unknown-id")
except NotFoundError:
    print("Analysis not found")
except QuotaError:
    print("Quota exceeded")
except AuthError:
    print("Invalid or expired API key")

Notebooks

Five Colab-ready notebooks are published with the developer documentation at https://www.dysplasiadx.com/developers. They are not installed by pip — this distribution is the client only.

RUO notice

Every scientific output carries ruo_disclaimer: "For Research Use Only. Not for use in diagnostic procedures.", and it is not suppressible by the caller.

That is the rule, not a list — it holds for analysis, cohort, counterfactual, interpretation, graph, experts, assay design, histology, radiomics, model builder and reports alike. The only exception is client.account, which returns administrative data (identity, credit balance, usage) and no scientific result.

DYSPLAI does not diagnose disease, recommend treatment, guide clinical management, provide clinical decision support, or generate prognosis.