Example notebooks demonstrating GPU-accelerated analysis of single-cell sequencing data. These notebooks use RAPIDS, which is a suite of open source software libraries and APIs providing the ability to execute end-to-end data science and analytics pipelines entirely on GPUs. Learn more about RAPIDS.
Examples are provided for both single-cell RNA-seq and single-cell ATAC-seq analysis. The workflow demonstrated in these notebooks begins with a count matrix that maps the level of activity of various genes or genomic elements in each cell. Subsequent steps include data preprocessing, machine learning, dimensionality reduction, clustering, visualization, and identification of genes or genomic elements that are differently active across clusters. A demonstration of GPU-accelerated analysis in an interactive cell browser is also provided.
These example notebooks are available in the clara-parabricks/rapids-single-cell-examples GitHub repository. The repository also contains equivalent CPU versions of the same notebooks for comparison.
Read this blog post for more information on the motivation behind the work and an overview of the workflow.