When doing bioinformatics analysis using a Python workstack, it is quite common that crucial software packages are implemented only in R.

This language divide often complicates reproducibility and development when trying to integrate R-based methods into Python-centric workflows. Technically, mamba, a next-gen reimplementation of conda in C++ can manage both Python and R ecosystems in a unified Conda environment. However, uv has quickly become the gold standard for managing Python projects thanks to its speed and precision, and it’s often good practice to isolate language-specific environments for better reproducibility and fewer conflicts.

In this post, we’ll walk through a clean setup for combining rpy2, uv, and mamba inside a Jupyter Notebook. This lets you run R code from Python seamlessly, leverage both ecosystems in a single workflow, and keep your environments modular and maintainable.

Step 1: Create a Conda Environment with R via Mamba

mamba create -n r45_env r-base=4.5 -y
mamba activate r45_env

Step 2: Set Up a uv Python Environment in Your Project Directory and Install Required Packages

uv venv
source .venv/bin/activate
uv add rpy2 ipykernel

Step 3: Register a Jupyter Kernel for This Hybrid Setup

uv run ipython kernel install --user --env VIRTUAL_ENV $(pwd)/.venv --name=uv-rpy2

Then edit the kernel JSON file (~/.local/share/jupyter/kernels/uv-rpy2/kernel.json) to point to the Conda R:

{
  "argv": [
    "/full/path/to/.venv/bin/python",
    "-m",
    "ipykernel_launcher",
    "-f",
    "{connection_file}"
  ],
  "display_name": "uv-rpy2",
  "language": "python",
  "env": {
    "R_HOME": "/opt/conda/envs/r45_env/lib/R",
    "LD_LIBRARY_PATH": "/opt/conda/envs/r45_env/lib/R/lib"
  }
}

Tip: You can find R_HOME by running R RHOME inside the activated mamba R environment.

Step 4: Verify It’s Working Inside Jupyter

In a notebook using the new kernel:

import rpy2.robjects as ro
print(ro.r("R.version.string")[0])

You should see 'R version 4.5.x (yyyy-mm-dd)'.