SOAPy_st.tl.cell_level_communications

SOAPy_st.tl.cell_level_communications(adata: AnnData, lr_pairs: lr_pairs, species: Literal['ligand', 'receptor'] = 'ligand', agg_method: Literal['min', 'mean', 'gmean'] = 'gmean', scale: str | float = 'hires', sample: str | int | list | None = None, sample_key: str | None = None, contact_key: str = 'contact', contact_max_quantile: int = 95, secretory_key: str = 'secretory', secretory_radius: int | float = 100, fdr_axis: Literal['spot', 'lr'] = 'spot', pval_adj_cutoff: float = 0.05, adj_method: str = 'fdr_bh', norm: bool = True, percent_of_drop: int | float = 0, spatial_in_obsm: str = 'spatial', n_iters: int = 1000, inplace: bool = True) → AnnData

A permutation test of ligand-receptor expression across every spot.

Parameters

adataanndata.AnnData

An AnnData object containing spatial omics data and spatial information.

lr_pairslr_pairs

database of ligand-receptor pairs.

speciesLiteral[‘ligand’, ‘receptor’], optional

‘ligand’: The central cell is the ligand cell and the neighbor cells are the receptor cells. ‘receptor’: The central cell is the receptor cell and the neighbor cells are the ligand cells.

agg_methodstr, optional

Integrated subunit to calculate ligand or receptor. ‘min’: ligand or receptor = min(sub1, sub2, sub3, …) ‘mean’: ligand or receptor = mean(sub1, sub2, sub3, …) ‘gmean’: geometrical mean, ligand or receptor = np.gmean(sub1, sub2, sub3, …)

scaleUnion[str, float], optional

scale used in subsequent analyses. If it’s Visium data it can also be HE image labels (hires or lower). Most of the time you don’t need to change this

sampleUnion[str, int, list], optional

Samples for which communication scores need to be calculated

sample_keystr, optional

The keyword of sample id in adata.obs.columns.

contact_keystr, optional

The tag name that represents the contact type in the LR database

contact_max_quantileint

In Neighbor network, Order the distance of all sides, and more than max_quantile% of the sides will be removed

secretory_keystr, optional

The tag name that represents the secretory type in the LR database

secretory_radiusUnion[int, float]

The maximum distance considered for the secretory type

fdr_axisLiteral[‘spot’, ‘lr’]

Dimensions of applying multiple hypothesis testing ‘spot’: tested in each spot ‘lr’: tested in each ligand-receptor pair None: no multiple hypothesis testing

pval_adj_cutofffloat

Cutoff for spot to be significant based on adjusted p-value.

adj_methodstr

Any method supported by statsmodels.stats.multitest.multipletests; https://www.statsmodels.org/dev/generated/statsmodels.stats.multitest.multipletests.html

normbool

Whether 0-1 normalization was performed on the expression volume data.

percent_of_dropUnion[int, float]

Percentage of extreme values removed from normalization

spatial_in_obsmstr

The key of spatial coordinates in adata.obsm

n_itersint

Number of permutations for the permutation-test

inplacebool, optional

Whether to change the original adata.

Returns

• anndata.AnnData.uns ['SOAPy']['ligand_spot_comm_score' or 'receptor_spot_comm_score']

  • ‘contact_names’: Names of used contact type lR pairs.

  • ‘secretory_names’: Names of used secretory type lR pairs.

  • ‘contact_affinity’: The affinity of contact type lR pairs.

  • ‘secretory_affinity’: The affinity of secretory type lR pairs.