Skip to Content
🧬 BioQuery is in beta. We'd love your feedback!
SdkPython SDK

Python SDK

The BioQuery Python SDK provides a simple, Pythonic interface for querying cancer genomics data.

Features

  • Natural Language Queries: Ask questions in plain English
  • Interactive Visualizations: Plotly figures you can customize
  • Statistical Analysis: Publication-ready p-values and effect sizes
  • Multiple Data Sources: TCGA, TARGET, GTEx, CCLE, CPTAC, GENIE
  • Export Options: PNG, SVG, JSON, DataFrame

Quick Example

import bioquery
 
# Initialize client
client = bioquery.Client(api_key="your-api-key")
 
# Ask a question in natural language
card = client.query("Is DDR1 expression higher in KIRP vs KIRC?")
 
# View the answer
print(card.answer)
# "DDR1 expression is significantly higher in KIRP compared to KIRC
#  (median 8.2 vs 5.1 TPM, p < 0.001, Wilcoxon rank-sum test)."
 
# Access statistical results
print(f"P-value: {card.p_value}")
print(f"Effect size: {card.effect_size}")
 
# Display interactive figure (in Jupyter)
card.show_figure()
 
# Export
card.save_figure("ddr1_comparison.png")
df = card.to_dataframe()

Use Cases

Research Workflows

Integrate BioQuery into your analysis pipelines:

genes = ["DDR1", "EGFR", "TP53", "BRCA1"]
results = []
 
for gene in genes:
    card = client.query(f"Is {gene} higher in KIRP vs KIRC?")
    results.append({
        "gene": gene,
        "p_value": card.p_value,
        "effect_size": card.effect_size,
    })
 
df = pd.DataFrame(results)

Jupyter Notebooks

Perfect for exploratory analysis:

# Interactive visualization in notebook
card = client.query("Show DDR1 expression across all TCGA cancers")
card.show_figure()
 
# Export for publication
card.save_figure("figure1.svg", format="svg")

Automated Reports

Generate reproducible reports:

card = client.query("Does high EGFR affect survival in lung adenocarcinoma?")
 
# Get grant-ready methods text
print(card.methods_text)
 
# Export full card for archival
with open("analysis.json", "w") as f:
    f.write(card.to_json())

Next Steps

Understanding ResultsClient Reference