Skip to main content
Start building
with Synthesize Bio
Generative models for human gene expression — predict bulk and single-cell profiles across tissues, cell types, perturbations, and disease states from metadata alone.
Available through Python and R SDKs, MCP-compatible tools, and the web.

Quickstart

API key

import pysynthbio

# Get an example query
query = pysynthbio.get_example_query(model_id="gem-1-bulk")["example_query"]

# Generate a dataset
result = pysynthbio.predict_query(query, model_id="gem-1-bulk")

# Access the results
metadata = result["metadata"]
expression = result["expression"]

Python logoPython SDK
Use pysynthbio to predict gene expression from your Python notebooks and pipelines. Jump to the Python SDK or start from the Quickstart.

R logoR SDK
Use rsynthbio to predict gene expression from your R notebooks and pipelines. Jump to the R SDK or start from the Quickstart.

MCP logoMCP
Use MCP-compatible AI tools to predict gene expression and chat about your results.

Synthesize Bio logoPlatform
Low- and no-code exploration in the browser — no local install required.

​
About our models

Generate synthetic gene expression profiles from metadata alone, then use them in downstream analysis and integration workflows.

GEM-1 bulk

Baseline model for predicting bulk gene expression.

GEM-1 single-cell

Baseline model for predicting single-cell gene expression.

GEM-2

Coming soon — next-generation capabilities. Check the platform and these docs for updates.

​
Advanced platform capabilities

Reference conditioning

Anchor generation to an existing, real reference sample while you apply perturbations or other modifications.

Metadata prediction

Predict or infer biological characteristics such as cell type, tissue, disease state, and more. Vocabularies and fields are model-specific — see Available metadata.
Reach out to partnerships@synthesize.bio if you are interested in trying out the advanced capabilities.

​
Support

Schedule a meeting

Email us to set up time with the team.

Email us

support@synthesize.bio for product and integration questions.