-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathpatient_database_analysis
202 lines (184 loc) · 6.16 KB
/
patient_database_analysis
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
import numpy as np
import pandas as pd
import seaborn as sns
from scipy.stats import ttest_ind
import statsmodels.sandbox.stats.multicomp as mcp
def Clean_data(data_file, label_file, label1, label2):
dat = pd.read_csv(data_file, sep = '\t')
dat = dat.T
x =dat.loc['Gene Symbol']
dat.columns = x
dat = dat.drop(dat.index[0:2])
dat_labels =pd.read_csv(label_file, header = None)
dat_labels = pd.Series(dat_labels[0])
dat_labels = dat_labels.tolist()
dat.index =dat_labels
dat = select_genes(dat, genes)
dat_met = dat.ix[label2]
dat_prim = dat.ix[label1]
dat, dat_sig = Ttest(dat_met, dat_prim)
return dat, dat_sig
def Ttest(data1, data2):
p_values =[]
fold_change = []
for i in data1:
try:
p = ttest_ind(data1[i], data2[i], nan_policy = 'omit')
p_values.append(p[1])
fold_change.append(data2[i].mean()/data1[i].mean())
print("gene:%s p= %.3f, fc =%.3f" %(i, p[1], data2[i].mean()/data1[i].mean()))
except TypeError:
print ("TypeError in gene %s, moving on" %i)
next
analysis = pd.DataFrame({'p_value':p_values,'fold_change':fold_change})
temp =mcp.multipletests(analysis['p_value'], alpha = 0.05, method = 'fdr_bh')
analysis['fdr_adjusted_pvalues'] = temp[1]
#analysis['log2_foldchange']=np.log2(analysis['fold_change'])
#analysis['log_qvalue'] = -(np.log(analysis['fdr_adjusted_pvalues']))
analysis.index = data1.columns
print("there are %i genes in the analysis" %(analysis.shape[0]))
analysis_sig = analysis[analysis.fdr_adjusted_pvalues < 0.05]
print("there are %i signficant genes" %(analysis_sig.shape[0]))
a1 = analysis_sig[analysis_sig.fold_change >1.5]
a2 = analysis_sig[analysis_sig.fold_change < (2.0/3.0)]
analysis_sig = pd.concat((a1,a2))
print("there are %i signficant genes with large fold change" %analysis_sig.shape[0])
return analysis, analysis_sig
def select_genes(df, genes):
x = []
for i in genes:
try:
#select primer with largest signal
try:
if df[i].shape[1]>1:
a = df[i].mean()
b = a.max()
d = df[i].ix[:,a==b]
x.append(d)
except IndexError:
x.append(df[i])
except KeyError:
print (i, " isn't found in the dataframe.")
next
x = pd.concat(x,1)
#get rid of duplicates
a1 = x.shape[1]
x = x.T.groupby(level =0).first().T
print("%i of %i genes found in the microarray" %(x.shape[1], len(genes)))
print("%i duplicate genes found" %(a1-x.shape[1]))
return x
def Z_score_gene(data):
x = data.mean()
std = data.std()
data = (x-data)/std
#get genes of interest
genes = pd.read_csv('new_genes.csv', header =None)
genes = genes.drop(0,1)
genes = np.array(genes)
genes = genes.reshape(genes.shape[0],)
print ("%i probes" %genes.shape[0])
genes = np.unique(genes)
print ("%i unique genes" %genes.shape[0])
genes = genes.tolist()
#clean up talentov
tal = pd.read_csv('Talentov_ready.txt', sep = '\t')
tal = tal.T
x = tal.loc['Gene Symbol']
tal.columns = x
tal =tal.drop(tal.index[[0,1]])
tal_index = pd.read_csv('tal_index.csv', header =None)
tal_index = pd.Series(tal_index[0])
tal_index = tal_index.tolist()
tal.index = tal_index
tal = select_genes(tal,genes)
nevus = tal.ix['Nevus']
melanoma = tal.ix['Melanoma']
tal, tal_sig = Ttest(nevus,melanoma)
tal.to_csv('talentov_all.csv')
tal_sig.to_csv('talentov_signficant.csv')
#clean up avery. it throws different datatype error
avery = pd.read_csv('Avery_ready1.csv')
avery = avery.T
avery.columns = avery.ix[1]
avery = avery.drop(avery.index[0:2])
av_labels = pd.read_csv('avery_labels.csv', header = None)
av_labels = pd.Series(av_labels[0])
av_labels = av_labels.tolist()
avery.index= av_labels
avery = select_genes(avery,genes)
#split data into conditions for analysis
melanocyte = avery.ix['Melanocyte']
avery_melanoma = avery.ix['Melanoma']
av, av_sig = Ttest(melanocyte, avery_melanoma)
av.to_csv('avery_all.csv')
av_sig.to_csv('avery_sign.csv')
#clean up kabbarah
kab = pd.read_csv('Hoe_ready.txt', sep = '\t')
kab = kab.T
x =kab.loc['Gene Symbol']
kab.columns = x
kab = kab.drop(kab.index[0:2])
kab_labels =pd.read_csv('kab_labels.csv', header = None)
kab_labels = pd.Series(kab_labels[0])
kab_labels = kab_labels.tolist()
kab.index =kab_labels
kab = select_genes(kab, genes)
kab_met = kab.ix['Metastatic Melanoma']
kab_prim = kab.ix['Primary Melanoma']
kab, kab_sig = Ttest(kab_met, kab_prim)
kab.to_csv('kabbarrah_all.csv')
kab_sig.to_csv('kab_sig.csv')
#clean up xu
xu = pd.read_csv('xu_ready.txt', sep = '\t')
xu = xu.T
x =xu.loc['Gene Symbol']
xu.columns = x
xu = xu.drop(xu.index[0:2])
xu_labels =pd.read_csv('xu_labels.csv', header = None)
xu_labels = pd.Series(xu_labels[0])
xu_labels = xu_labels.tolist()
xu.index =xu_labels
xu = select_genes(xu, genes)
xu_met = xu.ix['metastasis']
xu_prim = xu.ix['primary']
xu, xu_sig = Ttest(xu_met, xu_prim)
xu.to_csv('xu_all.csv')
xu_sig.to_csv('xu_sig.csv')
#Riker stuff
rik = pd.read_csv('Riker_ready.txt', sep = '\t')
rik = rik.T
x =rik.loc['Gene Symbol']
rik.columns = x
rik = rik.drop(rik.index[0:2])
rik_labels =pd.read_csv('rik_labels.csv', header = None)
rik_labels = pd.Series(rik_labels[0])
rik_labels = rik_labels.tolist()
rik.index =rik_labels
rik = select_genes(rik, genes)
rik_met = rik.ix['Metastatic Melanoma']
rik_prim = rik.ix['Primary Melanoma']
rik, rik_sig = Ttest(rik_met, rik_prim)
rik.to_csv('rik_all.csv')
rik_sig.to_csv('rik_sig.csv')
tcg = pd.read_csv('tcga_part1', sep = '\t')
tcg2 = pd.read_csv('tcga_part2', sep ='\t')
patients = pd.read_csv('tcga_patients.csv')
tcg=tcg.T
tcg2 = tcg2.T
t = pd.concat([tcg,tcg2], 1)
tcg = t
tcg.columns = tcg.ix['COMMON']
tcg = tcg.drop(tcg.index[0:2])
patients.index = patients.Patient
patients = patients.sort_index()
tcg = tcg.sort_index()
tcg = tcg.T
tcg = tcg.drop('Unnamed: 356',1)
tcg = tcg.T
tcg.index = patients.Location
tcg= tcg.apply(pd.to_numeric)
tcg =tcg.T.groupby(level=0).first().T
t_met =tcg.ix['metastasis']
t_prime = tcg.ix['primary tumour']
t_all, t_sig = Ttest(t_prime, t_met)
#y =y.T.groupby(level=0).first().T