3.1 CITE-seq GEX¶
In [1]:
from clustergrammer2 import net
df = {}
clustergrammer2 backend version 0.2.7
In [72]:
import matplotlib.pyplot as plt
%matplotlib inline
import json_scripts
In [2]:
import numpy as np
import pandas as pd
import gene_exp_10x
In [3]:
filename = '../data/CITE-seq_data/GSE100866_CBMC_8K_13AB_10X-RNA_umi.csv.gz'
df['gex-ini'] = pd.read_csv(filename, compression='gzip', index_col=0)
In [62]:
print(df['gex-ini'].shape)
(36280, 8617)
In [9]:
rows = df['gex-ini'].index.tolist()
len(rows)
Out[9]:
36280
In [10]:
mouse_genes = [x for x in rows if 'MOUSE_' in x]
len(mouse_genes)
Out[10]:
15879
In [11]:
human_genes = [x for x in rows if 'HUMAN_' in x]
len(human_genes)
Out[11]:
20400
In [12]:
len(mouse_genes) + len(human_genes)
Out[12]:
36279
In [21]:
print(len(mouse_genes))
print(len(list(set(mouse_genes))))
15879 15879
In [22]:
df['gex-mouse'] = df['gex-ini'].loc[mouse_genes]
df['gex-mouse'].shape
Out[22]:
(15879, 8617)
Filtering out mouse cells¶
In [47]:
ser_bc_mouse_sum = df['gex-mouse'].sum(axis=0).sort_values(ascending=False)
ser_bc_mouse_sum.plot('line', logy=True)
Out[47]:
<matplotlib.axes._subplots.AxesSubplot at 0x11321cda0>
In [48]:
ser_bc_mouse_sum[ser_bc_mouse_sum<125].shape
Out[48]:
(7688,)
In [49]:
ser_bc_mouse_sum[ser_bc_mouse_sum<100].shape
Out[49]:
(7339,)
In [50]:
ser_bc_mouse_sum[ser_bc_mouse_sum<50].shape
Out[50]:
(2317,)
Human cells appear to have <100 mouse gene counts¶
In [69]:
mouse_cells = ser_bc_mouse_sum[ser_bc_mouse_sum>=100].index.tolist()
human_cells = ser_bc_mouse_sum[ser_bc_mouse_sum<100].index.tolist()
In [70]:
print('number of mouse cells', len(mouse_cells))
print('number of human cells', len(human_cells))
number of mouse cells 1278 number of human cells 7339
~15% Mouse Cells¶
In [57]:
(8617-7339)/8617
Out[57]:
0.14831147731228966
Make DataFrame with Only Human Cells and Human Genes¶
In [61]:
df['gex-human'] = df['gex-ini'][human_cells].loc[human_genes]
df['gex-human'].shape
Out[61]:
(20400, 7339)
In [66]:
df['gex-human'].to_csv('../data/CITE-seq_data/GSE100866_CBMC_8K_13AB_10X-RNA_umi_HUMAN.csv.gz', compression='gzip')
In [67]:
filename = '../data/CITE-seq_data/GSE100866_CBMC_8K_13AB_10X-RNA_umi_HUMAN.csv.gz'
df['test'] = pd.read_csv(filename, compression='gzip', index_col=0)
In [68]:
df['test'].shape
Out[68]:
(20400, 7339)
Save dictionary of cell types¶
In [71]:
cell_dict = {}
cell_dict['mouse-cells'] = mouse_cells
cell_dict['human-cells'] = human_cells
In [73]:
json_scripts.save_to_json(filename='../data/CITE-seq_data/human_mouse_cell_dictionary.json', inst_dict=cell_dict)
In [ ]: