Skip to contents

Download and load the ircolitis tissue CD8 demo dataset

load_ircolitis_cd8.Rd

Downloads GSE206299_ircolitis-tissue-cd8.h5ad.gz from GEO on first call and caches it under tools::R_user_dir(). Subsequent calls reuse the cache. Used to demonstrate presto on a realistic single-cell dataset.

Usage

load_ircolitis_cd8(
  cache_dir = tools::R_user_dir("presto", which = "cache"),
  overwrite = FALSE,
  verbose = TRUE
)

Arguments

cache_dir

Directory for the cached .h5ad file. Defaults to tools::R_user_dir("presto", which = "cache").

overwrite

If TRUE, re-download even if the file is cached.

verbose

Print download / parsing progress.

Value

A list with elements:

  • counts - gene-by-cell dgCMatrix of raw UMI counts.

  • obs - cell metadata data.frame (categorical fields decoded to character; UMAP1/UMAP2 and PC1..PC24 kept as numeric).

  • var - gene metadata data.frame.

Details

Source: Thomas et al. (Nat. Med. 2024), GSE206299. 25,341 colon-tissue CD8 T cells from immune checkpoint colitis patients and controls.

Requires the suggested package rhdf5 (Bioconductor).

Examples

if (FALSE) { # \dontrun{
  d <- load_ircolitis_cd8()
  dim(d$counts)
  table(d$obs$cluster)
  res <- wilcoxauc(d$counts, d$obs$cluster)
} # }