This function will automatically generate a basic config file from a single-cell object. This can greatly speed up setup for large objects with many metadata variables, but we recommend editing the file in the config app or manually afterward for the following reasons:
All metadata variables will be added, in the order they appear in the object metadata table. A separate display order may be desired to highight the most important metadata variables to end users.
Display names for metadata are un-set
Display settings for modalities/assays, including how they appear in feature dropdown menus and whether to display a suffix after the assay name in plots, are unset.
Descriptions for individual modalities/assays, metadata, and reductions are unset. To load the config file in the config app for further editing, supply the file path to
config_pathinrun_config. After loading the config app, go to the "..." menu on the top right, and select "Load Config File". Save changes via "Save Config File" when you are finished editing.
Usage
auto_config(
# Replace with path to your object
object = path_to_your_object,
output_yaml = output_config_path,
object_name = "Test Single-Cell Object",
object_description = "This is a test object",
genes_assay = "RNA",
include_numeric_metadata = TRUE,
is_HDF5SummarizedExperiment = FALSE,
sample_level_var = NULL
)Arguments
- object
a single-cell object to be configured for use in scExploreR (not the path to the object). Currently, Seurat, SingleCellExperiment, and anndata objects are supported.
- output_yaml
path specifying where the output config file should be saved. This should end in ".yaml".
- object_name
a single-length character vector with the display name of the object. This will be displayed to end users in the app as entered.
- object_description
a single-length character vector with a short description of the object. This will be displayed to end users in the dataset selection menu. This should be a few sentences long, and include object details relevant to end-users (i.e. brief descripton of methods, number of cells/samples in object, etc.).
- is_HDF5SummarizedExperiment
Set this to TRUE to load an HDF5-enabled SingleCellExperiment object saved via saveHDF5SummarizedExperiment. When loading an HDF5-enabled object, set the object_path to the directory of the HDF5-enabled object, created when saving the object via HDF5Array:saveHDF5SummarizedExperiment.
- HDF5_prefix
When loading an HDF5-backed SingleCellExperiment HDF5_prefix is passed to the prefix parameter of HDF5Array::loadHDF5SummarizedExperiment to specify the prefixes for the se.rds and assays.h5 files.
- include_numeric_metadata
If
TRUE, numeric metadata in the object will searchable in scExploreR for plotting and subsetting. The default setting isTRUE.- genes_assay
single-length character vector giving the name of the genes assay/experiment/modality in the object. If undefined, the first assay in the object will be used. The name should be entered using the following convention:
Seurat objects: use the name of the assay as it appears in
names(object@assays).SingleCellExperiment objects: use the name of the experiment as it appears in
SingleCellExperiment::altExpNames()orSingleCellExperiment::mainExpName()anndata objects: use either "X" to specify the modality measured in X, or enter the name of a matrix stored in the
obsmslot.
- sample_level_var
a sample-level metadata variable used to construct sample-level pie charts in the plots tab. This should be a sample or patient ID, for example. If this is not provided (the default), pie charts will not appear in scExploreR.