cudacellr: GPU-Aware Single-Cell Analysis Workflows

Provides single-cell analysis primitives for sparse count matrices, including library-size normalization, log-normalization, highly variable feature selection, principal component analysis, and nearest-neighbour construction. Integrates the 'cudaverse' numerical and sparse layers while preserving matrix identifiers and recording the computation backend for each stage. Optional adapters preserve native 'SingleCellExperiment' and 'SeuratObject' version 5 object structures.

Version:	0.4.0
Imports:	cudaverse (≥ 0.4.1), Matrix, methods
Suggests:	DelayedArray, knitr, rmarkdown, S4Vectors, SeuratObject (≥ 5.0.0), SingleCellExperiment, SummarizedExperiment, testthat (≥ 3.0.0)
Published:	2026-10-27
DOI:	10.32614/CRAN.package.cudacellr (may not be active yet)
Author:	Yaoxiang Li [aut, cre]
Maintainer:	Yaoxiang Li <liyaoxiang at outlook.com>
BugReports:	https://github.com/cudaverse/cudacellr/issues
License:	MIT + file LICENSE
URL:	https://cudaverse.github.io/cudacellr/, https://github.com/cudaverse/cudacellr
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	cudacellr results

Documentation:

Reference manual:	cudacellr.html , cudacellr.pdf
Vignettes:	Backend provenance and CUDA diagnostics (source, R code) A native SeuratObject v5 workflow (source, R code) A native SingleCellExperiment workflow (source, R code)

Downloads:

Package source:	cudacellr_0.4.0.tar.gz
Windows binaries:	r-devel: not available, r-release: cudacellr_0.4.0.zip, r-oldrel: not available
macOS binaries:	r-release (arm64): cudacellr_0.4.0.tgz, r-oldrel (arm64): cudacellr_0.4.0.tgz, r-release (x86_64): cudacellr_0.4.0.tgz, r-oldrel (x86_64): cudacellr_0.4.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=cudacellr to link to this page.