README.md

In LTLA/CellTypeReferences: Index of Reference Cell Type Datasets

Index of reference cell type datasets for Bioconductor

|Environment|Status| |---|---| |BioC-release|| |BioC-devel||

This package provides reference datasets with annotated cell types for convenient use by other Bioconductor packages and workflows. References were originally sourced from the first edition of the SingleR package but more datasets have been added since then. Each dataset is loaded as a SummarizedExperiment that is immediately ready for further analysis. To get started, install the package and its dependencies from Bioconductor:

# install.packages("BiocManager")
BiocManager::install("celldex")

Find datasets of interest:

surveyReferences()
searchReferences(defineTextQuery("immune"))

Fetch a dataset as a SummarizedExperiment:

fetchReference("immgen", version="2024-02-26")
fetchReference("hpca", "2024-02-26")

And add your own datasets to enable re-use by the wider Bioconductor community.

Check out the user's guide for more details.

Maintainer notes

If someone wants to contribute a new dataset, they should follow the instructions in the user's guide. This requires a bit of effort on our (i.e., the package maintainers') part to process a new submission.

Once a prospective uploader has prepared their SummarizedExperiment and its associated metadata, they can be given temporary upload permissions for, e.g., a week, by calling:

gypsum::setPermissions("celldex", uploaders=list(
    list(
        id="GITHUB_LOGIN", 
        asset="NAME_OF_THE_DATASET_THEY_WANT_TO_UPLOAD",
        version="VERSION_THEY_WANT_TO_UPLOAD",
        until=Sys.time() + 7 * 24 * 3600
    )
)

Once the upload is complete, we pull down the dataset for review.

cache <- tempfile()
dest <- gypsum::saveVersion(
    "celldex", 
    asset="NAME_OF_THE_DATASET_THEY_WANT_TO_UPLOAD",
    version="VERSION_THEY_WANT_TO_UPLOAD",
    cache=cache
)

# Check that the saved object is valid. You might need to do this for each
# subdirectory if they saved multiple objects in a single dataset.
alabaster.base::validateObject(dest)

# You can also just try loading it for inspection in the R session.
alabaster.base::readObject(dest)

# Check that the metadata is valid.
lines <- readLines(file.path(dest, "_bioconductor.json"))
gypsum::validateMetadata(paste(lines, collapse="\n"))

If everything looks okay, we can approve the probational dataset. Otherwise we reject it.

# Okay.
gypsum::approveProbation(
    "celldex", 
    asset="NAME_OF_THE_DATASET_THEY_WANT_TO_UPLOAD",
    version="VERSION_THEY_WANT_TO_UPLOAD"
)

# Not okay.
gypsum::rejectProbation(
    "celldex", 
    asset="NAME_OF_THE_DATASET_THEY_WANT_TO_UPLOAD",
    version="VERSION_THEY_WANT_TO_UPLOAD"
)

LTLA/CellTypeReferences documentation built on June 1, 2024, 12:12 p.m.