RAPIDS is a suite of open-source Python libraries that can speed up data science workflows using GPU acceleration. Starting from a single-cell count matrix, RAPIDS libraries can be used to perform data processing, dimensionality reduction, clustering, visualization, and comparison of cell clusters.

Several of our examples are inspired by the Scanpy tutorials and based upon the AnnData format. Currently, we provide examples for scRNA-seq and scATAC-seq, and we have scaled up to 1 million cells. We also show how to create GPU-powered interactive, in-browser visualizations to explore single-cell datasets.

Dataset sizes for single-cell genomics studies are increasing, presently reaching millions of cells. With RAPIDS, it becomes easy to analyze large datasets interactively and in real time, enabling faster scientific discoveries.

Software Requirements

• Linux OS
• Pascal, Volta, Turing, or an NVIDIA Ampere architecture-based GPU.
• Nvidia Driver
• CUDA 11.4
• Docker
• Nvidia Docker container toolkit

Hardware Requirements

The sample notebooks are tested using A100 and V100 GPUs. Required number of GPUs depends on the dataset size. The largest dataset referred in the notebooks requires 32GB GPU memory.

• Two V100 or a A100.
• Hard disk space - 100GB
• System RAM - 64GB

Getting started

• Execute the following commands to start the container and follow the URL in the log to open Jupyter web application.

  DATADIR=<</host/directory/to/store/data>> docker run \ --gpus all \ --rm -it \ --network host \ -p 8888:8888 \ -v ${DATADIR}:/workspace/rapids-single-cell-examples/data \ nvcr.io/nvstaging/clara/single-cell-examples_rapids:v0.0.4 \ /opt/conda/envs/rapids/bin/jupyter-lab \ --no-browser \ --port=8888 \ --ip=0.0.0.0 \ --notebook-dir=/workspace/rapids-single-cell-examples/notebooks \ --NotebookApp.password="" \ --NotebookApp.token="" \ --NotebookApp.password_required=False \ --allow-root

• Select 'RAPIDS' kernel before executing the notebooks.