Read tabular and bioinformatics file formats

A Shennong wrapper around rio::import() with support for common single-cell and omics formats such as 10x directories, .h5, .h5ad, BPCells directories, and GMT files.

This adapter is exported so rio can dispatch to it.

This function imports Gene Matrix Transposed (GMT) files into a two-column data.frame of term–gene pairs.

Adapted from clusterProfiler::read.gmt() (Y. Yu et al., 2012–2024), which is licensed under the Artistic License 2.0.

Usage

sn_read(path, format, to = NULL, which, row_names = NULL, ...)

.import.rio_bpcells(file, ...)

.import.rio_10x(file, ...)

.import.rio_starsolo(file, ...)

.import.rio_h5ad(file, ...)

.import.rio_h5(file, ...)

.import.rio_gmt(file, ...)

Arguments

path: Path, URL, or supported directory to import.
format: Optional explicit format override.
to: Optional target class forwarded to rio.
which: Optional archive member index for compressed archives.
row_names: Optional column index or name to promote to row names when importing a data frame.
...: Additional arguments forwarded to the underlying importer.
file: Input path used by the exported rio adapter methods.

Value

The imported object. Tabular inputs are typically returned as data frames, while supported matrix-like formats are returned in their native classes.

A data.frame with two columns: term and gene.

Details

The original implementation is available in the clusterProfiler package: https://github.com/YuLab-SMU/clusterProfiler

Author

Yu Guangchuang (original implementation)

Examples

tmp <- tempfile(fileext = ".csv")
write.csv(mtcars[1:3, 1:3], tmp, row.names = FALSE)
sn_read(tmp)
#>    mpg cyl disp
#> 1 21.0   6  160
#> 2 21.0   6  160
#> 3 22.8   4  108