Cell-cell communication from single-cell dataset¶
In [1]:
Copied!
import cell2cell as c2c
import scanpy as sc
import pandas as pd
%matplotlib inline
import cell2cell as c2c
import scanpy as sc
import pandas as pd
%matplotlib inline
In [2]:
Copied!
import warnings
warnings.filterwarnings('ignore')
import warnings
warnings.filterwarnings('ignore')
Data¶
RNA-seq data¶
We begin by loading the single-cell transcriptomics data. For this tutorial, we will use a lung dataset of 63k immune and epithelial cells across three control, three moderate, and six severe COVID-19 patients21. We use a convenient function to download the data and store it in the AnnData format, on which the scanpy package is built.
In [3]:
Copied!
rnaseq = c2c.datasets.balf_covid()
rnaseq = c2c.datasets.balf_covid()
In [4]:
Copied!
rnaseq
rnaseq
Out[4]:
AnnData object with n_obs × n_vars = 63103 × 33538
obs: 'sample', 'sample_new', 'group', 'disease', 'hasnCoV', 'cluster', 'celltype', 'condition'
In [5]:
Copied!
rnaseq.obs.head()
rnaseq.obs.head()
Out[5]:
| sample | sample_new | group | disease | hasnCoV | cluster | celltype | condition | |
|---|---|---|---|---|---|---|---|---|
| AAACCCACAGCTACAT_3 | C100 | HC3 | HC | N | N | 27.0 | B | Control |
| AAACCCATCCACGGGT_3 | C100 | HC3 | HC | N | N | 23.0 | Macrophages | Control |
| AAACCCATCCCATTCG_3 | C100 | HC3 | HC | N | N | 6.0 | T | Control |
| AAACGAACAAACAGGC_3 | C100 | HC3 | HC | N | N | 10.0 | Macrophages | Control |
| AAACGAAGTCGCACAC_3 | C100 | HC3 | HC | N | N | 10.0 | Macrophages | Control |
For the purpose of this example, we will inspect only a patient with Severe COVID-19
In [6]:
Copied!
rnaseq = rnaseq[rnaseq.obs.sample_new == 'S1']
rnaseq = rnaseq[rnaseq.obs.sample_new == 'S1']
In [7]:
Copied!
rnaseq
rnaseq
Out[7]:
View of AnnData object with n_obs × n_vars = 11762 × 33538
obs: 'sample', 'sample_new', 'group', 'disease', 'hasnCoV', 'cluster', 'celltype', 'condition'
Then we do simple preprocessing of the data. For more standard data processing, refer to these best practices.
In [8]:
Copied!
sc.pp.filter_cells(rnaseq, min_genes=200)
sc.pp.filter_genes(rnaseq, min_cells=3)
sc.pp.filter_cells(rnaseq, min_genes=200)
sc.pp.filter_genes(rnaseq, min_cells=3)
Protein-Protein Interactions or Ligand-Receptor Pairs¶
In this case, we use a list of LR pairs published with CellChat (Jin et al. 2021, Nature Communications).
In [9]:
Copied!
lr_pairs = pd.read_csv('https://raw.githubusercontent.com/LewisLabUCSD/Ligand-Receptor-Pairs/master/Human/Human-2020-Jin-LR-pairs.csv')
lr_pairs = lr_pairs.astype(str)
lr_pairs = pd.read_csv('https://raw.githubusercontent.com/LewisLabUCSD/Ligand-Receptor-Pairs/master/Human/Human-2020-Jin-LR-pairs.csv')
lr_pairs = lr_pairs.astype(str)
In [10]:
Copied!
lr_pairs.head()
lr_pairs.head()
Out[10]:
| interaction_name | pathway_name | ligand | receptor | agonist | antagonist | co_A_receptor | co_I_receptor | evidence | annotation | interaction_name_2 | ligand_symbol | receptor_symbol | ligand_ensembl | receptor_ensembl | interaction_symbol | interaction_ensembl | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | TGFB1_TGFBR1_TGFBR2 | TGFb | TGFB1 | TGFbR1_R2 | TGFb agonist | TGFb antagonist | nan | TGFb inhibition receptor | KEGG: hsa04350 | Secreted Signaling | TGFB1 - (TGFBR1+TGFBR2) | TGFB1 | TGFBR1&TGFBR2 | ENSG00000105329 | ENSG00000106799&ENSG00000163513 | TGFB1^TGFBR1&TGFBR2 | ENSG00000105329^ENSG00000106799&ENSG00000163513 |
| 1 | TGFB2_TGFBR1_TGFBR2 | TGFb | TGFB2 | TGFbR1_R2 | TGFb agonist | TGFb antagonist | nan | TGFb inhibition receptor | KEGG: hsa04350 | Secreted Signaling | TGFB2 - (TGFBR1+TGFBR2) | TGFB2 | TGFBR1&TGFBR2 | ENSG00000092969 | ENSG00000106799&ENSG00000163513 | TGFB2^TGFBR1&TGFBR2 | ENSG00000092969^ENSG00000106799&ENSG00000163513 |
| 2 | TGFB3_TGFBR1_TGFBR2 | TGFb | TGFB3 | TGFbR1_R2 | TGFb agonist | TGFb antagonist | nan | TGFb inhibition receptor | KEGG: hsa04350 | Secreted Signaling | TGFB3 - (TGFBR1+TGFBR2) | TGFB3 | TGFBR1&TGFBR2 | ENSG00000119699 | ENSG00000106799&ENSG00000163513 | TGFB3^TGFBR1&TGFBR2 | ENSG00000119699^ENSG00000106799&ENSG00000163513 |
| 3 | TGFB1_ACVR1B_TGFBR2 | TGFb | TGFB1 | ACVR1B_TGFbR2 | TGFb agonist | TGFb antagonist | nan | TGFb inhibition receptor | PMID: 27449815 | Secreted Signaling | TGFB1 - (ACVR1B+TGFBR2) | TGFB1 | ACVR1B&TGFBR2 | ENSG00000105329 | ENSG00000135503&ENSG00000163513 | TGFB1^ACVR1B&TGFBR2 | ENSG00000105329^ENSG00000135503&ENSG00000163513 |
| 4 | TGFB1_ACVR1C_TGFBR2 | TGFb | TGFB1 | ACVR1C_TGFbR2 | TGFb agonist | TGFb antagonist | nan | TGFb inhibition receptor | PMID: 27449815 | Secreted Signaling | TGFB1 - (ACVR1C+TGFBR2) | TGFB1 | ACVR1C&TGFBR2 | ENSG00000105329 | ENSG00000123612&ENSG00000163513 | TGFB1^ACVR1C&TGFBR2 | ENSG00000105329^ENSG00000123612&ENSG00000163513 |
Metadata¶
Metadata for the single cells
In [11]:
Copied!
meta = rnaseq.obs.copy()
meta = rnaseq.obs.copy()
In [12]:
Copied!
meta.head()
meta.head()
Out[12]:
| sample | sample_new | group | disease | hasnCoV | cluster | celltype | condition | n_genes | |
|---|---|---|---|---|---|---|---|---|---|
| AAACCTGAGCCCGAAA_9 | C145 | S1 | S | Y | N | 15.0 | Neutrophil | Severe COVID-19 | 691 |
| AAACCTGAGCTACCGC_9 | C145 | S1 | S | Y | N | 0.0 | Macrophages | Severe COVID-19 | 744 |
| AAACCTGAGGAGTAGA_9 | C145 | S1 | S | Y | N | 4.0 | Macrophages | Severe COVID-19 | 2003 |
| AAACCTGAGGCCCGTT_9 | C145 | S1 | S | Y | N | 2.0 | Macrophages | Severe COVID-19 | 1702 |
| AAACCTGCAAATCCGT_9 | C145 | S1 | S | Y | N | 3.0 | Macrophages | Severe COVID-19 | 1254 |
Cell-cell Interactions and Communication Analysis¶
Using an Interaction Pipeline¶
The pipeline integrates the RNA-seq and PPI datasets by using the analysis setups. It generates an interaction space containing an instance for each sample/cell type, containing the values assigned to each protein in the PPI list given the setups for computing the CCI and CCC scores.
In this case is single cell data.
In [13]:
Copied!
interactions = c2c.analysis.SingleCellInteractions(rnaseq_data=rnaseq.to_df().T,
ppi_data=lr_pairs,
metadata=meta,
interaction_columns=('ligand_symbol', 'receptor_symbol'),
communication_score='expression_thresholding',
expression_threshold=0.1, # values after aggregation
cci_score='bray_curtis',
cci_type='undirected',
aggregation_method='nn_cell_fraction',
barcode_col='index',
celltype_col='celltype',
complex_sep='&',
verbose=False)
interactions = c2c.analysis.SingleCellInteractions(rnaseq_data=rnaseq.to_df().T,
ppi_data=lr_pairs,
metadata=meta,
interaction_columns=('ligand_symbol', 'receptor_symbol'),
communication_score='expression_thresholding',
expression_threshold=0.1, # values after aggregation
cci_score='bray_curtis',
cci_type='undirected',
aggregation_method='nn_cell_fraction',
barcode_col='index',
celltype_col='celltype',
complex_sep='&',
verbose=False)
Compute communication scores for each PPI or LR pair¶
In [14]:
Copied!
interactions.compute_pairwise_communication_scores()
interactions.compute_pairwise_communication_scores()
Computing pairwise communication Computing communication score between Epithelial and Epithelial Computing communication score between T and T Computing communication score between pDC and B Computing communication score between NK and B Computing communication score between Macrophages and B Computing communication score between Epithelial and T Computing communication score between Epithelial and pDC Computing communication score between Mast and T Computing communication score between NK and Epithelial Computing communication score between Neutrophil and Neutrophil Computing communication score between B and pDC Computing communication score between mDC and pDC Computing communication score between Macrophages and Epithelial Computing communication score between B and mDC Computing communication score between mDC and mDC Computing communication score between pDC and T Computing communication score between Plasma and NK Computing communication score between NK and T Computing communication score between Mast and Mast Computing communication score between Macrophages and pDC Computing communication score between Neutrophil and NK Computing communication score between Macrophages and mDC Computing communication score between pDC and Mast Computing communication score between Neutrophil and Macrophages Computing communication score between B and B Computing communication score between mDC and B Computing communication score between Plasma and pDC Computing communication score between B and Epithelial Computing communication score between mDC and Epithelial Computing communication score between Plasma and mDC Computing communication score between Mast and Neutrophil Computing communication score between Neutrophil and pDC Computing communication score between Neutrophil and mDC Computing communication score between B and T Computing communication score between T and Plasma Computing communication score between mDC and T Computing communication score between T and Mast Computing communication score between pDC and Neutrophil Computing communication score between Epithelial and Plasma Computing communication score between Mast and NK Computing communication score between Mast and Plasma Computing communication score between Epithelial and Mast Computing communication score between Macrophages and T Computing communication score between Plasma and B Computing communication score between Mast and Macrophages Computing communication score between pDC and Plasma Computing communication score between NK and Plasma Computing communication score between NK and Mast Computing communication score between Neutrophil and B Computing communication score between Plasma and Epithelial Computing communication score between pDC and Macrophages Computing communication score between Macrophages and Mast Computing communication score between T and Neutrophil Computing communication score between Neutrophil and Epithelial Computing communication score between Plasma and T Computing communication score between Epithelial and Neutrophil Computing communication score between Mast and mDC Computing communication score between Neutrophil and T Computing communication score between B and Neutrophil Computing communication score between T and NK Computing communication score between Plasma and Mast Computing communication score between NK and Neutrophil Computing communication score between T and Macrophages Computing communication score between Epithelial and NK Computing communication score between Macrophages and Neutrophil Computing communication score between B and Plasma Computing communication score between mDC and Plasma Computing communication score between Epithelial and Macrophages Computing communication score between B and Mast Computing communication score between mDC and Mast Computing communication score between Mast and B Computing communication score between pDC and NK Computing communication score between NK and NK Computing communication score between B and Macrophages Computing communication score between mDC and Macrophages Computing communication score between Macrophages and NK Computing communication score between Macrophages and Plasma Computing communication score between Mast and Epithelial Computing communication score between T and pDC Computing communication score between NK and Macrophages Computing communication score between Plasma and Neutrophil Computing communication score between T and mDC Computing communication score between Macrophages and Macrophages Computing communication score between pDC and Epithelial Computing communication score between Mast and pDC Computing communication score between Epithelial and mDC Computing communication score between mDC and Neutrophil Computing communication score between pDC and pDC Computing communication score between NK and pDC Computing communication score between Plasma and Plasma Computing communication score between pDC and mDC Computing communication score between NK and mDC Computing communication score between Plasma and Macrophages Computing communication score between Neutrophil and Plasma Computing communication score between T and B Computing communication score between Neutrophil and Mast Computing communication score between B and NK Computing communication score between mDC and NK Computing communication score between Epithelial and B Computing communication score between T and Epithelial
Compute CCI scores for each pair of cells¶
It is computed according to the analysis setups. We computed an undirected Bray-Curtis-like score for each pair, meaning that score(C1, C2) = score(C2, C1). Notice that our score is undirected, so for C1 and C2 was not necessary to compute C2 and C1 as happened for the communication scores
In [15]:
Copied!
interactions.compute_pairwise_cci_scores()
interactions.compute_pairwise_cci_scores()
Computing pairwise interactions Computing interaction score between B and B Computing interaction score between Macrophages and Mast Computing interaction score between Macrophages and Plasma Computing interaction score between T and pDC Computing interaction score between Epithelial and Epithelial Computing interaction score between T and T Computing interaction score between T and mDC Computing interaction score between Plasma and pDC Computing interaction score between Plasma and T Computing interaction score between Macrophages and Macrophages Computing interaction score between B and Epithelial Computing interaction score between Plasma and mDC Computing interaction score between Epithelial and T Computing interaction score between Mast and Neutrophil Computing interaction score between Mast and pDC Computing interaction score between Epithelial and Neutrophil Computing interaction score between Epithelial and pDC Computing interaction score between Mast and T Computing interaction score between Mast and mDC Computing interaction score between Neutrophil and pDC Computing interaction score between Epithelial and mDC Computing interaction score between Neutrophil and T Computing interaction score between Neutrophil and Neutrophil Computing interaction score between B and Neutrophil Computing interaction score between B and pDC Computing interaction score between Neutrophil and mDC Computing interaction score between B and T Computing interaction score between mDC and pDC Computing interaction score between B and mDC Computing interaction score between pDC and pDC Computing interaction score between mDC and mDC Computing interaction score between NK and pDC Computing interaction score between Plasma and Plasma Computing interaction score between NK and T Computing interaction score between NK and Neutrophil Computing interaction score between NK and mDC Computing interaction score between Epithelial and Plasma Computing interaction score between Mast and NK Computing interaction score between Epithelial and NK Computing interaction score between Mast and Plasma Computing interaction score between Mast and Mast Computing interaction score between Epithelial and Mast Computing interaction score between Macrophages and Neutrophil Computing interaction score between Macrophages and pDC Computing interaction score between Neutrophil and Plasma Computing interaction score between Macrophages and T Computing interaction score between B and Plasma Computing interaction score between Macrophages and mDC Computing interaction score between B and NK Computing interaction score between Epithelial and Macrophages Computing interaction score between B and Mast Computing interaction score between NK and Plasma Computing interaction score between NK and NK Computing interaction score between B and Macrophages Computing interaction score between Macrophages and NK
Perform permutation analysis¶
Just for demonstration, we are performing only 10 permutations here. We recommend using at least 100, ideally 1000. In addition, enable fdr correction with fdr_correction=True.
In [16]:
Copied!
cci_pvals = interactions.permute_cell_labels(evaluation='interactions',
permutations=10,
fdr_correction=False,
verbose=True)
cci_pvals = interactions.permute_cell_labels(evaluation='interactions',
permutations=10,
fdr_correction=False,
verbose=True)
100%|███████████████████████████████████████████████████████████████| 10/10 [00:24<00:00, 2.45s/it]
In [17]:
Copied!
cci_pvals
cci_pvals
Out[17]:
| B | Epithelial | Macrophages | Mast | NK | Neutrophil | Plasma | T | mDC | pDC | |
|---|---|---|---|---|---|---|---|---|---|---|
| B | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| Epithelial | 0.181818 | 0.363636 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| Macrophages | 0.181818 | 0.181818 | 0.090909 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| Mast | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| NK | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| Neutrophil | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| Plasma | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| T | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| mDC | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 |
| pDC | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.181818 | 0.363636 |
In [18]:
Copied!
ccc_pvals = interactions.permute_cell_labels(evaluation='communication',
permutations=10,
fdr_correction=False,
verbose=True)
ccc_pvals = interactions.permute_cell_labels(evaluation='communication',
permutations=10,
fdr_correction=False,
verbose=True)
100%|███████████████████████████████████████████████████████████████| 10/10 [00:25<00:00, 2.52s/it]
In [19]:
Copied!
ccc_pvals
ccc_pvals
Out[19]:
| Epithelial;Epithelial | T;T | pDC;B | NK;B | Macrophages;B | Epithelial;T | Epithelial;pDC | Mast;T | NK;Epithelial | Neutrophil;Neutrophil | ... | pDC;mDC | NK;mDC | Plasma;Macrophages | Neutrophil;Plasma | T;B | Neutrophil;Mast | B;NK | mDC;NK | Epithelial;B | T;Epithelial | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (TGFB1, TGFBR1&TGFBR2) | 1.000000 | 0.909091 | 0.181818 | 0.181818 | 0.181818 | 0.909091 | 0.909091 | 0.909091 | 1.000000 | 0.909091 | ... | 0.090909 | 0.090909 | 0.909091 | 0.909091 | 0.181818 | 0.181818 | 0.909091 | 0.909091 | 0.181818 | 1.000000 |
| (TGFB2, TGFBR1&TGFBR2) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| (TGFB3, TGFBR1&TGFBR2) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| (TGFB1, ACVR1B&TGFBR2) | 0.090909 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.090909 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.363636 | 0.909091 | 0.909091 | 0.909091 | 0.090909 |
| (TGFB1, ACVR1C&TGFBR2) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| (SIGLEC1, SPN) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.090909 | 0.090909 | 0.909091 | 0.909091 | 0.909091 | ... | 0.090909 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.727273 | 0.909091 | 0.090909 | 0.909091 | 0.909091 |
| (ITGA4&ITGB1, VCAM1) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| (ITGA9&ITGB1, VCAM1) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| (ITGA4&ITGB7, VCAM1) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
| (VSIR, IGSF11) | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | ... | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 | 0.909091 |
1006 rows × 100 columns
Visualizations¶
If we want to save the figure as a vector figure, we can pass a filename input to each plot function to save the figure. E.g. filename='/Users/cell2cell/CommScores.svg'
Generate a metadata for the cell types
In [20]:
Copied!
meta.celltype.unique()
meta.celltype.unique()
Out[20]:
['Neutrophil', 'Macrophages', 'T', 'NK', 'Mast', 'pDC', 'mDC', 'Epithelial', 'Plasma', 'B'] Categories (10, object): ['B', 'Epithelial', 'Macrophages', 'Mast', ..., 'Plasma', 'T', 'mDC', 'pDC']
In [21]:
Copied!
group_meta = pd.DataFrame(columns=['Celltype', 'Group'])
group_meta['Celltype'] = meta.celltype.unique().tolist()
group_meta['Group'] = ['Myeloid', 'Myeloid', 'Lymphoid', 'Lymphoid', 'Myeloid', 'Myeloid', 'Myeloid', 'Epithelial',
'Lymphoid', 'Lymphoid']
group_meta = pd.DataFrame(columns=['Celltype', 'Group'])
group_meta['Celltype'] = meta.celltype.unique().tolist()
group_meta['Group'] = ['Myeloid', 'Myeloid', 'Lymphoid', 'Lymphoid', 'Myeloid', 'Myeloid', 'Myeloid', 'Epithelial',
'Lymphoid', 'Lymphoid']
In [22]:
Copied!
group_meta
group_meta
Out[22]:
| Celltype | Group | |
|---|---|---|
| 0 | Neutrophil | Myeloid |
| 1 | Macrophages | Myeloid |
| 2 | T | Lymphoid |
| 3 | NK | Lymphoid |
| 4 | Mast | Myeloid |
| 5 | pDC | Myeloid |
| 6 | mDC | Myeloid |
| 7 | Epithelial | Epithelial |
| 8 | Plasma | Lymphoid |
| 9 | B | Lymphoid |
Generate colors for the groups in the metadata
It returns a dictionary with the colors.
In [23]:
Copied!
colors = c2c.plotting.get_colors_from_labels(labels=group_meta['Group'].unique().tolist(),
cmap='tab10'
)
colors = c2c.plotting.get_colors_from_labels(labels=group_meta['Group'].unique().tolist(),
cmap='tab10'
)
Visualize communication scores for each LR pair and each cell pair¶
In [24]:
Copied!
interaction_clustermap = c2c.plotting.clustermap_ccc(interactions,
metric='jaccard',
method='complete',
metadata=group_meta,
sample_col='Celltype',
group_col='Group',
colors=colors,
row_fontsize=14,
title='Active ligand-receptor pairs for interacting cells',
filename=None,
cell_labels=('SENDER-CELLS', 'RECEIVER-CELLS'),
**{'figsize' : (10,9),
}
)
# Add a legend to know the groups of the sender and receiver cells:
l1 = c2c.plotting.generate_legend(color_dict=colors,
loc='center left',
bbox_to_anchor=(20, -2), # Indicated where to include it
ncol=1, fancybox=True,
shadow=True,
title='Groups',
fontsize=14,
)
interaction_clustermap = c2c.plotting.clustermap_ccc(interactions,
metric='jaccard',
method='complete',
metadata=group_meta,
sample_col='Celltype',
group_col='Group',
colors=colors,
row_fontsize=14,
title='Active ligand-receptor pairs for interacting cells',
filename=None,
cell_labels=('SENDER-CELLS', 'RECEIVER-CELLS'),
**{'figsize' : (10,9),
}
)
# Add a legend to know the groups of the sender and receiver cells:
l1 = c2c.plotting.generate_legend(color_dict=colors,
loc='center left',
bbox_to_anchor=(20, -2), # Indicated where to include it
ncol=1, fancybox=True,
shadow=True,
title='Groups',
fontsize=14,
)
Interaction space detected as an InteractionSpace class
Circos plot¶
It generates a circos plot, showing the cells producing ligands (sender_cells), according to a list of specific ligands (ligands) to the cells producing receptors (receiver_cells), according to a list of specific receptors (receptors). The order of these lists is preserved in the visualization. Elements that are not used are omitted and therefore not plotted (in this case those with a communication score of 0, specified in 'excluded_score').
In [25]:
Copied!
sender_cells = ['Epithelial', 'pDC', 'mDC']
receiver_cells = ['Neutrophil', 'NK', 'T', 'B', 'Macrophages']
ligands = ['CCL5',
'MIF',
'LAMB2'
]
receptors = ['NCL',
'CD74&CD44',
'CCR1',
]
sender_cells = ['Epithelial', 'pDC', 'mDC']
receiver_cells = ['Neutrophil', 'NK', 'T', 'B', 'Macrophages']
ligands = ['CCL5',
'MIF',
'LAMB2'
]
receptors = ['NCL',
'CD74&CD44',
'CCR1',
]
In [26]:
Copied!
c2c.plotting.circos_plot(interaction_space=interactions,
sender_cells=sender_cells,
receiver_cells=receiver_cells,
ligands=ligands,
receptors=receptors,
excluded_score=0,
metadata=group_meta,
sample_col='Celltype',
group_col='Group',
colors=colors,
fontsize=15,
)
c2c.plotting.circos_plot(interaction_space=interactions,
sender_cells=sender_cells,
receiver_cells=receiver_cells,
ligands=ligands,
receptors=receptors,
excluded_score=0,
metadata=group_meta,
sample_col='Celltype',
group_col='Group',
colors=colors,
fontsize=15,
)
Out[26]:
<Axes: >
If we do not pass metadata info, the samples/cell types are only plotted.
We can also change the label colors of ligands and receptors
In [27]:
Copied!
c2c.plotting.circos_plot(interaction_space=interactions,
sender_cells=sender_cells,
receiver_cells=receiver_cells,
ligands=ligands,
receptors=receptors,
excluded_score=0,
fontsize=20,
ligand_label_color='orange',
receptor_label_color='brown',
)
c2c.plotting.circos_plot(interaction_space=interactions,
sender_cells=sender_cells,
receiver_cells=receiver_cells,
ligands=ligands,
receptors=receptors,
excluded_score=0,
fontsize=20,
ligand_label_color='orange',
receptor_label_color='brown',
)
Out[27]:
<Axes: >
Dot plot for significance¶
In [28]:
Copied!
fig = c2c.plotting.dot_plot(interactions,
evaluation='communication',
significance = 0.2,
figsize=(9, 16),
cmap='PuOr',
senders=sender_cells,
receivers=receiver_cells,
tick_size=8
)
fig = c2c.plotting.dot_plot(interactions,
evaluation='communication',
significance = 0.2,
figsize=(9, 16),
cmap='PuOr',
senders=sender_cells,
receivers=receiver_cells,
tick_size=8
)
Visualize CCI scores¶
Since this case is undirected, a triangular heatmap is plotted instead of the complete one.
In [29]:
Copied!
cm = c2c.plotting.clustermap_cci(interactions,
method='complete',
metadata=group_meta,
sample_col="Celltype",
group_col="Group",
colors=colors,
title='CCI scores for cell-types',
cmap='Blues'
)
# Add a legend to know the groups of the sender and receiver cells:
l1 = c2c.plotting.generate_legend(color_dict=colors,
loc='center left',
bbox_to_anchor=(20, -2), # Indicated where to include it
ncol=1, fancybox=True,
shadow=True,
title='Groups',
fontsize=14,
)
cm = c2c.plotting.clustermap_cci(interactions,
method='complete',
metadata=group_meta,
sample_col="Celltype",
group_col="Group",
colors=colors,
title='CCI scores for cell-types',
cmap='Blues'
)
# Add a legend to know the groups of the sender and receiver cells:
l1 = c2c.plotting.generate_legend(color_dict=colors,
loc='center left',
bbox_to_anchor=(20, -2), # Indicated where to include it
ncol=1, fancybox=True,
shadow=True,
title='Groups',
fontsize=14,
)
Interaction space detected as a Interactions class
Dot plot for significance¶
In [30]:
Copied!
fig = c2c.plotting.dot_plot(interactions,
evaluation='interactions',
significance = 0.2,
figsize=(16, 9),
cmap='Blues',
)
fig = c2c.plotting.dot_plot(interactions,
evaluation='interactions',
significance = 0.2,
figsize=(16, 9),
cmap='Blues',
)
Project samples/cells with PCoA into a Euclidean space¶
We can project the samples/cells given their CCI scores with other cells and see how close they are given their potential of interaction.
THIS ONLY WORKS WITH UNDIRECTED CCI SCORES
In [31]:
Copied!
if interactions.analysis_setup['cci_type'] == 'undirected':
pcoa = c2c.plotting.pcoa_3dplot(interactions,
metadata=group_meta,
sample_col="Celltype",
group_col="Group",
title='PCoA based on potential CCIs',
colors=colors,
)
if interactions.analysis_setup['cci_type'] == 'undirected':
pcoa = c2c.plotting.pcoa_3dplot(interactions,
metadata=group_meta,
sample_col="Celltype",
group_col="Group",
title='PCoA based on potential CCIs',
colors=colors,
)
Interaction space detected as a Interactions class
Without metadata
In [32]:
Copied!
if interactions.analysis_setup['cci_type'] == 'undirected':
celltype_colors = c2c.plotting.get_colors_from_labels(labels=meta['celltype'].unique().tolist(),
cmap='tab10'
)
pcoa = c2c.plotting.pcoa_3dplot(interactions,
title='PCoA based on potential CCIs',
colors=celltype_colors,
)
if interactions.analysis_setup['cci_type'] == 'undirected':
celltype_colors = c2c.plotting.get_colors_from_labels(labels=meta['celltype'].unique().tolist(),
cmap='tab10'
)
pcoa = c2c.plotting.pcoa_3dplot(interactions,
title='PCoA based on potential CCIs',
colors=celltype_colors,
)
Interaction space detected as a Interactions class
In [ ]:
Copied!