-
Notifications
You must be signed in to change notification settings - Fork 0
/
ex4_preprocess_lupus.py
196 lines (154 loc) · 5.72 KB
/
ex4_preprocess_lupus.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
import os
import sys
os.environ["OMP_NUM_THREADS"] = "11"
os.environ["OPENBLAS_NUM_THREADS"] = "8" # export OPENBLAS_NUM_THREADS=4
os.environ["MKL_NUM_THREADS"] = "11" # export MKL_NUM_THREADS=6
os.environ["VECLIB_MAXIMUM_THREADS"] = "8" # export VECLIB_MAXIMUM_THREADS=4
os.environ["NUMEXPR_NUM_THREADS"] = "11" # export NUMEXPR_NUM_THREADS=6
os.environ["NUMBA_CACHE_DIR"]='/tmp/numba_cache'
import numpy as np
import pandas as pd
import scipy as sp
import scipy.stats
import anndata as ad
import scanpy as sc
from sklearn.preprocessing import OrdinalEncoder, OneHotEncoder
########################################################################
#
# Load raw data
#
########################################################################
adata = sc.read_h5ad('data/lupus/GSE174188_CLUES1_adjusted.h5ad')
# only keep ential info to save memory
X = adata.raw.X
obs = pd.DataFrame()
obs['SLE_status'] = adata.obs['SLE_status'].tolist()
obs['ind_cov'] = adata.obs['ind_cov'].tolist()
obs['batch_cov'] = adata.obs['batch_cov'].tolist()
obs['pop_cov'] = adata.obs['pop_cov'].tolist()
obs['cg_cov'] = adata.obs['cg_cov'].tolist()
obs['ct_cov'] = adata.obs['ct_cov'].tolist()
obs['Age'] = adata.obs['Age'].tolist()
obs['Sex'] = adata.obs['Sex'].tolist()
obs['Processing_Cohort'] = adata.obs['Processing_Cohort'].tolist()
obs['L3'] = adata.obs['L3'].tolist()
var_names = adata.raw.var_names.tolist()
var = pd.DataFrame(index=var_names)
cdata = ad.AnnData(X, obs=obs, var=var, dtype='int32')
cdata.raw = cdata
del adata
cdata.write_h5ad("data/lupus/raw.h5ad")
########################################################################
#
# Preprocess raw data
#
########################################################################
adata = sc.read_h5ad('data/lupus/raw.h5ad')
print(adata)
# AnnData object with n_obs × n_vars = 1263676 × 32738
# obs: 'SLE_status', 'ind_cov', 'batch_cov', 'pop_cov', 'cg_cov', 'ct_cov', 'Age', 'Sex', 'Processing_Cohort', 'L3'
adata.obs.groupby(['ind_cov']).count().shape[0]
# 261 individuals
adata = adata[adata.obs['pop_cov'].isin(['Asian', 'European']),:]
# 256 individuals
adata.obs['cg_pop_cov'] = adata.obs['cg_cov'].astype(str) + '-' + adata.obs['pop_cov'].astype(str)
adata.obs['cg_pop_cov'] = adata.obs['cg_pop_cov'].astype("category")
sc.pp.normalize_total(adata, target_sum=1e4)
sc.pp.log1p(adata)
adata.obs.groupby(['ind_cov']).count().shape[0]
# 256
########################################################################
#
# Save data with 2000 highly variable genes
#
########################################################################
col_covariate = ['Sex','Processing_Cohort','pop_cov','SLE_status']
path_data = 'data/lupus/'
for celltype in adata.obs['cg_cov'].unique()[:5]:
idx = adata.obs['cg_cov'] == celltype
_adata = adata[idx].copy()
sc.pp.highly_variable_genes(_adata, n_top_genes=2000)
gene_names = _adata.var.index[_adata.var['highly_variable']]
count = _adata.raw.X[:,np.in1d(_adata.raw.var_names.tolist(), gene_names)]
_df = _adata.obs
uni_ind = np.unique(_df['ind_cov'])
print(celltype)
print('\t shape:', _adata.raw.X.shape, len(_adata.raw.var_names))
print('\t ind:', len(uni_ind))
arr = []
meta = []
for ind in uni_ind:
arr.append(count[_df['ind_cov']==ind].sum(axis=0))
meta.append(_df[_df['ind_cov']==ind].iloc[:1,:])
Y = np.concatenate(arr, axis=0)
meta = pd.concat(meta, axis=0).reset_index(drop=True)
enc = OrdinalEncoder()
X = enc.fit_transform(meta[['Sex','pop_cov','SLE_status']])
enc = OneHotEncoder(drop='first')
_X = enc.fit_transform(meta[['Processing_Cohort']])
X = np.c_[X[:,:-1], _X.toarray(), X[:,-1:]]
file_name = path_data + 'data_lupus_{}.npz'.format(celltype)
with open(file_name, 'wb') as f:
np.savez(f, X=X, Y=Y, gene_names=gene_names)
# T4
# shape: (375885, 32738) 32738
# ind: 256
# cM
# shape: (302649, 32738) 32738
# ind: 256
# B
# shape: (149861, 32738) 32738
# ind: 254
# T8
# shape: (242677, 32738) 32738
# ind: 256
# NK
# shape: (90190, 32738) 32738
# ind: 256
########################################################################
#
# Save data with 250 highly variable genes
#
########################################################################
n_hvg = 250
for celltype in adata.obs['cg_cov'].unique()[:5]:
idx = adata.obs['cg_cov'] == celltype
_adata = adata[idx].copy()
sc.pp.highly_variable_genes(_adata, n_top_genes=n_hvg)
gene_names = _adata.var.index[_adata.var['highly_variable']]
count = _adata.raw.X[:,np.in1d(_adata.raw.var_names.tolist(), gene_names)]
_df = _adata.obs
uni_ind = np.unique(_df['ind_cov'])
print(celltype)
print('\t shape:', _adata.raw.X.shape, len(_adata.raw.var_names))
print('\t ind:', len(uni_ind))
arr = []
meta = []
for ind in uni_ind:
arr.append(count[_df['ind_cov']==ind].sum(axis=0))
meta.append(_df[_df['ind_cov']==ind].iloc[:1,:])
Y = np.concatenate(arr, axis=0)
meta = pd.concat(meta, axis=0).reset_index(drop=True)
enc = OrdinalEncoder()
X = enc.fit_transform(meta[['Sex','pop_cov','SLE_status']])
enc = OneHotEncoder(drop='first')
_X = enc.fit_transform(meta[['Processing_Cohort']])
X = np.c_[X[:,:-1], _X.toarray(), X[:,-1:]]
file_name = path_data + 'data_lupus_{}_{}.npz'.format(celltype,n_hvg)
with open(file_name, 'wb') as f:
np.savez(f, X=X, Y=Y, gene_names=gene_names)
# T4
# shape: (375885, 32738) 32738
# ind: 256
# cM
# shape: (302649, 32738) 32738
# ind: 256
# B
# shape: (149861, 32738) 32738
# ind: 254
# T8
# shape: (242677, 32738) 32738
# ind: 256
# NK
# shape: (90190, 32738) 32738
# ind: 256