API¶

crosschat.CrossChatH module¶

class crosschat.CrossChatH.CrossChatH(adata, species='human', user_comm_ids=None)¶

Bases: object

Cluster_LRs(LR_ls)¶

Cluster specific ligand-receptor pairs

Params: LR_ls is the list of ligand-receptor pairs to be clustered
Returns: clustering of ligand-receptor pairs

Cluster_pathways(nclusters=3)¶

Cluster active CCC pathways

Params: nclusters is the desired number of clusters of pathways
Returns: visualiztion of active CCC pathways after clustering

Detect_active_pathways()¶

Detect active CCC pathways

Returns: active pathways involved in CCC

Detect_specific_LRs(topN=20)¶

Detect specific ligands and receptors

Params: topN is the desired number of specific ligand-receptor pairs
Returns: list of specific ligand-receptor pairs

Draw_CCC(pathway, CCC_threshold=0.4, save=None)¶

Draw CCC between hierarchical clusters

Params: pathway is the pathway to visualize
Params: CCC_threshold is the threshold of CCC strengths to be visualized
Returns: visualization of hierarchical CCC

Draw_CCC_LR(ligand, receptor, CCC_threshold=0.4)¶

Draw_CCC_LR draws the CCC between a pair of ligand and receptor between ligand clustering and receptor clustering

Params: ligand/receptor is a list of ligands/receptors
Params: CCC_threshold is the threshold of CCC, interactions with strength below it are filtered
Returns: visualization of CCC

Draw_annotations_umap()¶

Draw umap of cell type annotations

Returns: multiscale umap

Draw_multiscale_umap(cluster_input='allgenes', spatial=False, save=None)¶

Draw umap of hierarchical clustering

Params: cluster_input is the input user wants to use for drawing umap. It can be “allgenes”,”L”,”R”, or “userinput”
Returns: multiscale umap

Multsicale_clustering(cluster_by='allgenes', k=15)¶

Runs multiscale clustering on cells based on either allgenes, or ligands, or receptors

Params: cluster_by is the genes that user want to use to cluster, either “allgenes” or “lr”
Params: k: value of k in knn
Returns: CrossChatH object after multiscale clustering

Multsicale_clustering_spatial(cluster_by='allgenes', k=15, w=0.5)¶

Runs multiscale clustering on spatial data based on either allgenes, or ligands, or receptors

Params: cluster_by is the genes that user want to use to cluster, either “allgenes” or “lr”
Params: k is value of k in knn
Params: w is the weight
Returns: CrossChatH object after multiscale clustering

jaccard_dist(comm_ids, celltype_annotations, save=False)¶

prepare_adata(normalize=False, scale=False, input='allgenes')¶

Prepares the CrossChatH object

Params: normalize is True if data needs to be normalized
Params: scale is True if data needs to be scaled
Params: input is allgenes if use all genes
Returns: the CrossChatH object

select_partitions(max_nvi=0.1, window_size=15, basin_radius=15, lr='L')¶

Select the desired hierarchical clustering

Params: lr is “L” (ligand) or “R” (receptor)
Returns: CrossChatH object after selecting partitions

crosschat.CrossChatT module¶

class crosschat.CrossChatT.CrossChatT(adata, species='human')¶

Bases: object

Binarization(threshold=0)¶

Binarizes the input gene expression matrix

Params: threshold is the cutoff percentage to throw out the cells with lowest gene expression according to the threshold
Returns: the CrossChatT object after binarization

Detect_trees(type='l', remove_cells_prop=0.9, support_size_threshold=30, inclusive_threshold=0.9, disjoint_threshold=0.95, tree_size=4, tree_scales=3)¶

Detect trees involved in CCC in scRNA-seq data.

Params: type is either “l” (ligand) or “r” (receptor)
Params: remove_cells_prop keeps genes that are present in more than the proportion of cells
Params: support_size_threshold keeps genes that are present in more than support_size_threshold cells
Params: tree_size is the threshold of number of nodes in the tree
Params: tree_scales is the number of levels in the tree
Returns: the CrossChatT object with detected trees

Detect_trees_S(spatial_range=300, type='l', remove_cells_prop=0.9, support_size_threshold=30, inclusive_threshold=0.9, disjoint_threshold=0.95, tree_size=4, tree_scales=3)¶

Detect trees involved in CCC in spatial data.

Params: type is either “l” (ligand) or “r” (receptor)
Params: remove_cells_prop keeps genes that are present in more than the proportion of cells
Params: support_size_threshold keeps genes that are present in more than support_size_threshold cells
Params: tree_size is the threshold of number of nodes in the tree
Params: tree_scales is the number of levels in the tree
Returns: the CrossChatT object with detected trees

Draw_CCC_between_MMT(lr_tree_inds)¶

Visualize CCC detected between trees

Params: lr_tree_inds is the index of ligand/receptor trees
Returns: the visualization of CCC between detected trees

Draw_MMT(type='l', tree_inds=None, nodesize=20)¶

Visualize detected trees

Params: type is either “l” or “r”
Params: tree_inds is the index of tree in all detected trees
Returns: the visualization of detected trees

Draw_MMT_lr_union(tree_inds=None)¶

Visualize detected trees with lr_union as input

Params: tree_inds is the index of tree in all detected trees
Returns: the visualization of detected trees

Draw_annotations_umap()¶

Draw umap of cell type annotations

Returns: multiscale umap

Draw_big_tree(type='l', tree_inds=None)¶

Draw union of trees

Params: type is the type of ligands or receptors
Params: tree_inds is the list of indices for the set of trees user is interested in
Returns: the visualization of frequency of ligand/receptor occurrence in ligands/receptors trees

find_interacting_trees(num_ligand_trees=10, num_receptor_trees=10, num_interation_threshold=1)¶

Obtain the pairs of interacting trees

Params: num_ligand_trees is the number of top detected ligand trees that user is interested
Params: num_receptor_trees is the number of top detected receptor trees that user is interested
Params: number_interaction_threshold is the threshold of CCC strength
Returns: the list of interacting trees

plot_lr_frequency(type='l')¶

Plot frequency of ligand/receptor occurrence in ligands/receptors trees

Params: type is the type of ligands or receptors
Returns: the visualization of frequency of ligand/receptor occurrence in ligands/receptors trees

prepare_adata(normalize=False, scale=False, input='allgenes')¶

Prepares the CrossChatT object.

Params: normalize is True if data needs to be normalized
Params: scale is True if data needs to be scaled
Params: input is allgenes if use all genes
Returns: the CrossChatH object

API¶

crosschat.CrossChatH module¶

crosschat.CrossChatT module¶

CrossChat

Navigation

Related Topics