spide

spide(
    adata,
    condition=None,
    sigma=None,
    engine='slopes',
    procedure=None,
    fdr=0.05,
    cell_type='cell_type',
    sample_id='sample_id',
    name='Niche',
    n_jobs=1,
    verbose=True,
    **fit_kwargs,
)

Run spiDE end to end (R spiDE).

Builds the niche covariates at bandwidth sigma (unless adata already carries them), fits the per-patient niche models (fit_spide) and tests them (test_spide): the pooled niche test always, and the condition-specific test when a condition is given.

Parameters

Name Type Description Default
adata AnnData Cells x genes with raw counts. Niches built here are not left on adata. required
condition str The obs column of a two-level, patient-level condition. None
sigma float The niche bandwidth (a single value). Required: None raises an error. None
engine ('slopes', 'sandwich') See fit_spide. "slopes"
procedure str Passed to test_spide; defaults to the engine’s default. None
fdr float Passed to test_spide. 0.05
cell_type str The obs column of the cell type. "cell_type"
sample_id str The obs column of the patient. "sample_id"
name str The niche key prefix: the niches are read from, or built into, adata.obsm[name + sigma]. "Niche"
n_jobs int Number of workers. 1
verbose bool Report progress on stderr. True
**fit_kwargs Further arguments to fit_spide. {}

Returns

Name Type Description
SpiDEResults The tests of the fit.

See Also

build_niches : The first step. fit_spide : The second step. test_spide : The third step.

Examples

from spidepy import spide
from spidepy.datasets import toy_spide
res = spide(toy_spide(), condition="condition", sigma=30, verbose=False)
res.fit.engine
'slopes'
res.results().columns.tolist()[:4]
['gene', 'index', 'niche', 'test']