metachat.pp.LRC_unfiltered

metachat.pp.LRC_unfiltered(adata, LRC_name=None, LRC_source='marker', obs_name=None, quantile=90.0, copy=False)[source]

Identify unfiltered candidate LRC (long-range channel) spots based on the quantile of a marker feature.

This function selects candidate points whose marker feature (e.g., gene expression or score) exceeds a specified quantile threshold. The result is stored in adata.obs['LRC_<LRC_name>_<LRC_source>_unfiltered'] as categorical values (0 or 1).

Parameters

adataanndata.AnnData: Annotated data matrix with shape n_obs × n_var.
LRC_namestr: The name of the long-range channel (e.g., 'Blood' or 'CSF').
LRC_sourcestr, default=’marker’: The type of feature used for selection (e.g., 'marker', 'score'). This will be included in the generated column name.
obs_namestr: The key in adata.obs containing the numeric feature used for quantile selection.
quantilefloat, default=90.0: The percentile threshold (0–100). Example: 90.0 means select all points above the 90th percentile.
copybool, default=False: If True, returns a copy of the modified AnnData object. Otherwise modifies the input object in place and returns None.

Returns

adataanndata.AnnData or None: If copy=True, returns a copy of the AnnData with a new column 'LRC_<LRC_name>_<LRC_source>_unfiltered' in .obs. Otherwise, modifies in place and returns None.

Notes

The resulting column is stored as a pandas Categorical with values {0, 1}.