The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

rsynthbio

rsynthbio is an R package that provides a convenient interface to the Synthesize Bio API, allowing users to generate realistic gene expression data based on specified biological conditions. This package enables researchers to easily access AI-generated transcriptomic data for various modalities, including bulk RNA-seq and single-cell RNA-seq.

To generate datasets without code, use our web platform.

Get started | Full R SDK docs

For function-level reference, use R’s built-in help (?predict_query, help(package = "rsynthbio")) or the CRAN reference manual.

For questions, suggestions, and support, email us at support@synthesize.bio.

How to install

You can install rsynthbio from CRAN:

install.packages("rsynthbio")

To install the development version from GitHub:

remotes::install_github("synthesizebio/rsynthbio")

Authentication

Set up your API token (obtain one at app.synthesize.bio):

library(rsynthbio)

# Securely prompt for your token (recommended)
set_synthesize_token()

# Or store in system keyring for persistence across sessions
set_synthesize_token(use_keyring = TRUE)

# Load from keyring in future sessions
load_synthesize_token_from_keyring()

Quick start

library(rsynthbio)

# See available models
list_models()

# Get an example query for bulk RNA-seq
query <- get_example_query(model_id = "gem-1-bulk")$example_query

# Generate synthetic expression data
result <- predict_query(query, model_id = "gem-1-bulk")

# Access the results
metadata <- result$metadata
expression <- result$expression

Available models

Model type	Bulk	Single-cell	Description
Baseline	`gem-1-bulk`	`gem-1-sc`	Generate expression from metadata
Reference conditioning	`gem-1-bulk_reference-conditioning`	`gem-1-sc_reference-conditioning`	Generate expression anchored to a reference sample
Metadata prediction	`gem-1-bulk_predict-metadata`	`gem-1-sc_predict-metadata`	Predict metadata from expression

Only baseline models are available to all users. Check programmatically with list_models(). Contact us at support@synthesize.bio if you have any questions.

Self-hosted models

Partners who run Synthesize models inside their own environment can point the same client at a self-hosted container. Self-hosted mode sends a single synchronous request and decodes the Apache Arrow IPC stream response into the same data frames as the hosted path (no polling, no download URL). It requires the optional arrow package and does not require an API key.

Self-hosted deployment is a model deployment option available within a Synthesize Bio partnership. To learn more or request access, contact partnerships@synthesize.bio.

install.packages("arrow")

# Point each model at its own container once (per-model environment variables)
Sys.setenv(
  SYNTHESIZE_SELF_HOSTED = "1",
  SYNTHESIZE_API_BASE_URL__GEM_1_BULK = "https://gem-1-bulk.internal.example",
  SYNTHESIZE_API_BASE_URL__GEM_1_SC   = "https://gem-1-sc.internal.example"
)

query <- get_example_query("gem-1-bulk", self_hosted = TRUE)$example_query
result <- predict_query(query, model_id = "gem-1-bulk", self_hosted = TRUE)
result$expression

You can also pass api_base_url explicitly per call instead of using the environment variables. Base-URL resolution precedence is: explicit api_base_url -> per-model SYNTHESIZE_API_BASE_URL__<MODEL> -> global SYNTHESIZE_API_BASE_URL -> production default. If the container is started with authentication enabled, set SYNTHESIZE_API_KEY and the client will send it as a bearer token. See the Self-Hosted Models vignette for details.

Documentation

For detailed usage and guides, see the R SDK section of the Synthesize Bio docs:

Getting started — Installation, authentication, and overview
Available metadata — Metadata fields you can query
Baseline models — Generate expression from metadata
Reference conditioning — Condition on real expression data
Metadata prediction — Infer metadata from expression
Self-hosted models — Run models in your own environment via Arrow streaming
Function reference — All exported functions

The legacy pkgdown site at synthesizebio.github.io/rsynthbio redirects to the canonical docs above; see docs-redirect/ and the deploy-docs-redirect workflow for how those redirects are served.

Authoring the Mintlify docs

The shared docs site at docs.synthesize.bio aggregates the R SDK section directly from this repo into /rsynthbio/... — there is no separate docs repo to update.

Authoring sources:

Long-form pages: vignettes/*.Rmd (also shipped as R vignettes for offline vignette(package = "rsynthbio") browsing)
Function reference: man/*.Rd (generated from R/*.R roxygen comments)
Generation script: scripts/generate_mintlify_docs.py

Regenerate the Mintlify output in docs-external/ after changing vignettes, roxygen docs, or the package surface:

python3 scripts/generate_mintlify_docs.py

The generated pages are committed in docs-external/ so changes are reviewable in PRs.

Rate limits

Free usage of Synthesize Bio is limited. If you exceed this limit, you’ll receive an error message. To generate more samples, please contact us at support@synthesize.bio.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.