-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathrun_data.py
213 lines (172 loc) · 8.64 KB
/
run_data.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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
from optimize import *
from optimize_fraction import *
import pandas as pd
from zipfile import ZipFile
import json
import gzip
import pickle
directory=Path('/data/liquid_biopsy/MOONSHOT2/bootstrap')
# apply the real data to method
# step1. generate the input format
''' gene_list: a list of total sequenced genes
tree_list: a list of dict for tree structure
node_list: a list of dict for tree node - mutation_list
clonal_freq_list: a list of dicts of the frequency for each node
tree_freq_list: a list of the frequencies of each possible tree
n_markers: number for the gene markers
'''
patient_num = 256
method = 'phylowgs'
tree_distribution_file = directory / f'{patient_num}_bootstrap_{method}/{method}_bootstrap_aggregation.pkl'
with open(directory / tree_distribution_file, 'rb') as f:
tree_distribution = pickle.load(f)
gene_list = []
gene2idx = {}
inter = pd.read_excel(directory / f"xlsx/Patient_{patient_num}_bootstrap.xlsx", sheet_name='common_blood_tissue_no_germline', index_col=0)
inter = inter[inter["Allele Frequency_x"] < 0.9]
inter = inter[inter["Allele Frequency_y"] < 0.9]
calls = inter
gene2idx = {'s' + str(i): i for i in range(len(inter))}
gene_list = list(gene2idx.keys())
gene_name_list = []
gene_count = {}
for i in range(inter.shape[0]):
gene = calls.iloc[i, 0]
if gene in gene_name_list:
gene_count[gene] += 1
gene = gene + '_' + str(gene_count[gene])
else:
gene_count[gene] = 1
if not isinstance(gene, str):
gene = calls.iloc[i, 1] + '_' + str(calls.iloc[i, 2])
gene_name_list.append(gene)
tree_list, node_list, clonal_freq_list, tree_freq_list = tree_distribution['tree_structure'], tree_distribution['node_dict'],tree_distribution['vaf_frac'],tree_distribution['freq']
#scrub node_list
node_list_scrub = []
for node_dict in node_list:
temp = {}
for key, values in node_dict.items():
temp.setdefault(int(key), values)
node_list_scrub.append(temp)
clonal_freq_list_scrub = []
for clonal_freq_dict in clonal_freq_list:
temp = {}
for key, values in clonal_freq_dict.items():
temp.setdefault(int(key), values[0])
clonal_freq_list_scrub.append(temp)
# tree_list_sub = [tree_list[i] for i in range(5) ]
# node_list_sub = [node_list_scrub[i] for i in range(5) ]
# clonal_freq_list_sub = [clonal_freq_list_scrub[i] for i in range(5) ]
# tree_freq_list_sub = [tree_freq_list[i] for i in range(5)]
selected_markers1_genename_ordered = []
obj1_ordered = []
for n_markers in range(1, len(gene_name_list) + 1):
#selected_markers1, obj = select_markers_fractions_weighted_single(gene_list, n_markers, tree_list, node_list_scrub, clonal_freq_list_scrub, gene2idx, tree_freq_list, 0)
selected_markers1, obj = select_markers_fractions_weighted_overall(gene_list, n_markers, tree_list, node_list_scrub, clonal_freq_list_scrub, gene2idx, tree_freq_list)
selected_markers1_genename = [gene_name_list[int(i[1:])] for i in selected_markers1]
obj1_ordered.append(obj)
if len(selected_markers1_genename) == 1:
selected_markers1_genename_ordered.append(selected_markers1_genename[0])
else:
diff_set = set(selected_markers1_genename).difference(set(selected_markers1_genename_ordered))
selected_markers1_genename_ordered.append(list(diff_set)[0])
print(selected_markers1_genename_ordered, obj1_ordered)
position1 = list(range(len(obj1_ordered)))
plt.figure(figsize=(8, 5))
plt.plot(position1, obj1_ordered, 'o-', label='tracing-fractions')
plt.xticks(position1, selected_markers1_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_tracing_subclones.eps', format='eps')
plt.show()
read_depth=1500
lam1=1 # lam1 is tree fractions
lam2=0 # lam2 is tree distributions
selected_markers2_genename_ordered = []
obj2_ordered = []
for n_markers in range(1, len(gene_name_list) + 1):
selected_markers2, obj_frac, obj_struct = select_markers_tree_gp(gene_list, n_markers, tree_list, node_list_scrub, clonal_freq_list_scrub, gene2idx, tree_freq_list, read_depth=read_depth, lam1=lam1, lam2=lam2)
selected_markers2_genename = [gene_name_list[int(i[1:])] for i in selected_markers2]
obj2_ordered.append((obj_frac, obj_struct))
if len(selected_markers2_genename) == 1:
selected_markers2_genename_ordered.append(selected_markers2_genename[0])
else:
selected_markers2_genename_ordered.append(
list(set(selected_markers2_genename).difference(set(selected_markers2_genename_ordered)))[0])
print(selected_markers2_genename_ordered, obj2_ordered)
import seaborn as sns
obj2_frac_ordered = [obj2_ordered[i][0] for i in range(len(obj2_ordered))]
obj2_struct_ordered = [obj2_ordered[i][1] for i in range(len(obj2_ordered))]
position2 = list(range(len(obj2_ordered)))
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_frac_ordered, 'o-', color='tab:orange',label='trees-fractions')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_fractions_{lam1}_{lam2}.eps', format='eps')
plt.show()
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_struct_ordered, 'o-', color='tab:green',label='trees-structure')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_structures_{lam1}_{lam2}.eps', format='eps')
plt.show()
read_depth=1500
lam1=0 # lam1 is tree fractions
lam2=1 # lam2 is tree distributions
selected_markers2_genename_ordered = []
obj2_ordered = []
for n_markers in range(1, len(gene_name_list) + 1):
selected_markers2, obj_frac, obj_struct = select_markers_tree_gp(gene_list, n_markers, tree_list, node_list_scrub, clonal_freq_list_scrub, gene2idx, tree_freq_list, read_depth=read_depth, lam1=lam1, lam2=lam2)
selected_markers2_genename = [gene_name_list[int(i[1:])] for i in selected_markers2]
obj2_ordered.append((obj_frac, obj_struct))
if len(selected_markers2_genename) == 1:
selected_markers2_genename_ordered.append(selected_markers2_genename[0])
else:
selected_markers2_genename_ordered.append(
list(set(selected_markers2_genename).difference(set(selected_markers2_genename_ordered)))[0])
print(selected_markers2_genename_ordered, obj2_ordered)
import seaborn as sns
obj2_frac_ordered = [obj2_ordered[i][0] for i in range(len(obj2_ordered))]
obj2_struct_ordered = [obj2_ordered[i][1] for i in range(len(obj2_ordered))]
position2 = list(range(len(obj2_ordered)))
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_frac_ordered, 'o-', color='tab:orange',label='trees-fractions')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_fractions_{lam1}_{lam2}_{read_depth}.eps', format='eps')
plt.show()
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_struct_ordered, 'o-', color='tab:green',label='trees-structure')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_structures_{lam1}_{lam2}_{read_depth}.eps', format='eps')
plt.show()
lam1=1 # lam1 is tree fractions
lam2=0 # lam2 is tree distributions
selected_markers2_genename_ordered = []
obj2_ordered = []
for n_markers in range(1, len(gene_name_list) + 1):
selected_markers2, obj_frac, obj_struct = select_markers_tree_gp(gene_list, n_markers, tree_list, node_list_scrub, clonal_freq_list_scrub, gene2idx, tree_freq_list, read_depth=read_depth, lam1=lam1, lam2=lam2)
selected_markers2_genename = [gene_name_list[int(i[1:])] for i in selected_markers2]
obj2_ordered.append((obj_frac, obj_struct))
if len(selected_markers2_genename) == 1:
selected_markers2_genename_ordered.append(selected_markers2_genename[0])
else:
selected_markers2_genename_ordered.append(
list(set(selected_markers2_genename).difference(set(selected_markers2_genename_ordered)))[0])
print(selected_markers2_genename_ordered, obj2_ordered)
import seaborn as sns
obj2_frac_ordered = [obj2_ordered[i][0] for i in range(len(obj2_ordered))]
obj2_struct_ordered = [obj2_ordered[i][1] for i in range(len(obj2_ordered))]
position2 = list(range(len(obj2_ordered)))
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_frac_ordered, 'o-', color='tab:orange',label='trees-fractions')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_fractions_{lam1}_{lam2}_{read_depth}.eps', format='eps')
plt.show()
plt.figure(figsize=(8, 5))
plt.plot(position2, obj2_struct_ordered, 'o-', color='tab:green',label='trees-structure')
plt.xticks(position2, selected_markers2_genename_ordered, rotation=30)
plt.legend()
plt.savefig(f'{patient_num}_trees_structures_{lam1}_{lam2}_{read_depth}.eps', format='eps')
plt.show()