-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy path9c_Model_specificity.py
187 lines (153 loc) · 8.24 KB
/
9c_Model_specificity.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
#!/usr/bin/env python3
import pandas as pd
import numpy as np
from sklearn.model_selection import StratifiedKFold
from sklearn.ensemble import RandomForestClassifier,GradientBoostingClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.svm import SVC,LinearSVC
from sklearn.metrics import roc_curve,auc,recall_score,precision_score,f1_score,accuracy_score,roc_auc_score
from numpy import interp
import matplotlib.pyplot as plt
import argparse
import seaborn as sns
from scipy.stats import wilcoxon
from mpl_toolkits.axes_grid1 import ImageGrid
plt.rcParams['pdf.fonttype'] = 42
plt.rcParams['ps.fonttype'] = 42
#import data
parser = argparse.ArgumentParser(description = "Model specificity assessment & plot")
parser.add_argument('--Workplace','-W',help = 'Workplace : Input and output work place')
parser.add_argument('--metadata','-m',help = 'input file : training set metadata')
parser.add_argument('--profile','-p',help = 'input file : optimal biomarkers')
parser.add_argument('--external_metadata','-q',help = 'input file : test set metadata')
parser.add_argument('--external_profile','-l',help = 'input file : test set microbial profile')
parser.add_argument('--other_metadata','-a',help = 'input file : test set metadata')
parser.add_argument('--other_profile','-x',help = 'input file : test set microbial profile')
parser.add_argument('--exposure','-e',help = 'input param : the control group name')
parser.add_argument('--group','-g',help = 'input param : the column name of experimental interest(group) in metadata')
parser.add_argument('--batch','-b',help = 'input param : the column name of cohort(dataset)')
parser.add_argument('--classifier','-c',help = 'input param : selected classifier')
parser.add_argument('--hyperparameter','-r',help = 'input param : tuned hyperparameters')
parser.add_argument('--number','-n',help = 'input param : number of samples for validation each time')
parser.add_argument('--seed','-s',help = 'input param : random seed')
parser.add_argument('--output','-o',help = 'output file prefix: External test result & plot ')
args = parser.parse_args()
metadata = pd.read_table(args.Workplace+args.metadata,sep = '\t',index_col = 0)
opt_biomarker = pd.read_table(args.Workplace+args.profile,sep = '\t',index_col=0)
data_group = np.array([0 if i== str(args.exposure) else 1 for i in metadata[str(args.group)]])
ex_metadata = pd.read_table(args.Workplace+args.external_metadata,sep = '\t',index_col = 0)
ex_data = pd.read_table(args.Workplace+args.external_profile,sep = '\t',index_col=0)
ex_data_group = np.array([0 if i== str(args.exposure) else 1 for i in ex_metadata[str(args.group)]])
ex_data = pd.DataFrame(ex_data,columns=opt_biomarker.columns)
ex_data = ex_data.fillna(0)
number = int(args.number)
other_metadata = pd.read_table(args.Workplace+args.other_metadata,sep = '\t',index_col = 0)
other_data = pd.read_table(args.Workplace+args.other_profile,sep = '\t',index_col=0)
other_data = pd.DataFrame(other_data,columns=opt_biomarker.columns)
other_data = other_data.fillna(0)
RANDOM_SEED = int(args.seed)
opt_clf = args.classifier
params = {}
file = open(args.Workplace+args.hyperparameter,'r')
for line in file.readlines():
line = line.strip()
k = line.split(' ')[0]
v = line.split(' ')[1]
params[k] = v
file.close()
best_param= [{k: int(v) if v and '.' not in v else float(v) if v else None for k, v in d.items()}for d in [params]][0]
#def function
class machine_learning:
def __init__(self):
self.Method = {'LRl1':LogisticRegression(penalty='l1', random_state=RANDOM_SEED, solver='liblinear', class_weight='balanced'),
'LRl2':LogisticRegression(penalty='l2', random_state=RANDOM_SEED, solver='liblinear', class_weight='balanced'),
'DT':DecisionTreeClassifier(class_weight='balanced', random_state=RANDOM_SEED),
'RF':RandomForestClassifier(oob_score=True, class_weight='balanced', random_state=RANDOM_SEED),
'GB':GradientBoostingClassifier(random_state=RANDOM_SEED),
'KNN':KNeighborsClassifier(n_neighbors=3),
'SVC':SVC(class_weight='balanced',random_state=RANDOM_SEED,probability = True)
}
def test_model(self,data, data_group, ex_data,ex_data_group,params):
clf = self.Method[opt_clf].set_params(**params).fit(opt_biomarker,data_group)
probas = clf.predict_proba(ex_data)
pred = clf.predict(ex_data)
sen = recall_score(ex_data_group,pred)
TP = ((pred==1) & (ex_data_group==1)).sum()
FP = ((pred==1) & (ex_data_group==0)).sum()
TN = ((pred==0) & (ex_data_group==0)).sum()
FN = ((pred==0) & (ex_data_group==1)).sum()
spe = TN / float(FP + TN)
pre = precision_score(ex_data_group, pred)
f1 = f1_score(ex_data_group, pred)
fpr, tpr, thresholds = roc_curve(ex_data_group, probas[:, 1])
roc_auc = auc(fpr, tpr)
accu = accuracy_score(ex_data_group, pred)
return clf, roc_auc
ML = machine_learning()
#Specificity assessment
cases = set(other_metadata[args.group])
cases.remove(str(args.exposure))
column_name = []
for i in cases:
column_name.append(i+"_control")
column_name.append(i+"_case")
auc_comparison = pd.DataFrame(columns = column_name,index = range(1,11,1))
ls=[]
ls.extend("0"*number)
ex_data_group_add = np.append(ex_data_group,ls).astype(int)
for i in cases:
print(i+" testing")
temp_set = list(set(other_metadata[other_metadata[args.group]==i][args.batch]))
temp_meta = other_metadata[other_metadata[args.batch].isin(temp_set)]
temp_control = other_data[other_data.index.isin(temp_meta[temp_meta[args.group]==args.exposure].index)]
temp_case = other_data[other_data.index.isin(temp_meta[temp_meta[args.group]!=args.exposure].index)]
for j in range (1,11,1):
add_control = temp_control.sample(n=number,axis=0,random_state=j)
ex_add_control = pd.concat((ex_data,temp_control.sample(n=number,axis=0,random_state=j)),axis=0)
ex_add_case = pd.concat((ex_data,temp_case.sample(n=number,axis=0,random_state=j)),axis=0)
_, roc_auc_control= ML.test_model(opt_biomarker,data_group,ex_add_control,ex_data_group_add, best_param)
_, roc_auc_case= ML.test_model(opt_biomarker,data_group,ex_add_case,ex_data_group_add, best_param)
auc_comparison.at[j,i+"_control"]=roc_auc_control
auc_comparison.at[j,i+"_case"]=roc_auc_case
auc_comparison.to_csv(args.Workplace+args.output+"_specificity_add_result.txt", sep = '\t')
#fig = plt.figure(figsize=(8,6))
#fig = sns.set_theme(style="white")
#fig = sns.boxplot(data = auc_comparison)
#fig = sns.swarmplot(data = auc_comparison)
#fig.set_ylabel('AUC')
#fig.set_xticklabels(auc_comparison.columns,rotation=30)
#fig.set_title("Model specificity assessment", fontsize=20, fontweight='bold', pad=20)
p_values = []
pairs = list(range(0, len(auc_comparison.columns) - 1, 2)) # [0, 2, 4, ...]
for i in pairs:
_, p_val = wilcoxon(auc_comparison.iloc[:, i], auc_comparison.iloc[:, i+1])
p_values.append(p_val)
significance_markers = []
for p in p_values:
if p < 0.001:
significance_markers.append('***')
elif p < 0.01:
significance_markers.append('**')
elif p < 0.05:
significance_markers.append('*')
else:
significance_markers.append('ns') # 不显著
sns.set_theme(style="white")
fig, ax = plt.subplots(figsize=(10, 6))
sns.boxplot(data=auc_comparison, ax=ax)
sns.swarmplot(data=auc_comparison, ax=ax, color=".25")
ax.set_ylabel('AUC')
ax.set_xticklabels(auc_comparison.columns)
ax.set_title("Biomarker specificity assessment", fontsize=20, fontweight='bold', pad=20)
y_max = auc_comparison.max().max()
for idx, marker in enumerate(significance_markers):
y_text = y_max + 0.02 #+ idx*0.02
x1, x2 = pairs[idx], pairs[idx] + 1
ax.plot([x1, x1, x2, x2], [y_text-0.01, y_text, y_text, y_text-0.01], lw=1.5, color='black')
ax.text((x1+x2)*.5, y_text, marker, ha='center', va='center', color='black', fontsize=15)
#plt.tight_layout()
plt.savefig(args.Workplace+args.output+'_specificity_add_auc.pdf',bbox_inches = 'tight')
plt.savefig(args.Workplace+args.output+'_specificity_add_auc.svg',bbox_inches = 'tight',format = 'svg')
print("FINISH")