-
Notifications
You must be signed in to change notification settings - Fork 1
/
Create_3_13_13.py
251 lines (190 loc) · 12.1 KB
/
Create_3_13_13.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
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
'''
Created on 11 Oct 2016
@author: sxj307
'''
from os import listdir
from Target_CCMPRED_MATRIX_3_13_13 import Target
import numpy as np
import os
root='' #input directory
outputroot='' #output directory
for domain in listdir(root):
profile_file =root+domain+'/'+'BestSequence.colstats'
pairwise_file =root+domain+'/'+'BestSequence.pairstats'
psicov_file =root+domain+'/'+'BestSequence.quic'
evfold_file =root+domain+'/'+'BestSequence.evfold'
ccmpred_file =root+domain+'/'+'BestSequence.ccmpred'
seconds_file =root+domain+'/'+'BestSequence.ss.spd3'
solvent_file =root+domain+'/'+'BestSequence.solv.spd3'
if not os.path.exists(profile_file): # This step ensures that all the files that are required are present
print "no profile"
continue
if not os.path.exists(pairwise_file):
print "no pairwise file"
continue
if not os.path.exists(psicov_file):
print "no psicov file"
continue
if not os.path.exists(evfold_file):
print "evfold file"
continue
if not os.path.exists(ccmpred_file):
print "ccmpred file"
continue
if not os.path.exists(seconds_file):
print "ss file"
continue
if not os.path.exists(solvent_file):
print "solvent file"
continue
outputhandle=open(outputroot+domain+'_BestSequence.metapsicov.features','wb')
target=Target()
with open(profile_file,'rU') as handle: #This step takes the info from colstats e.g sequence length, neff
target.seqlen=int(handle.readline()) # seqlen
target.nseq=int(handle.readline()) # number of sequences
target.effnseq=float(handle.readline()) # number of effective sequences
target.aacomposition=(handle.readline()).split() # amino acid composition
target.aacomposition=[float(i) for i in target.aacomposition]
line=handle.readline() #Each line is read and the the first 21 columns are added target.profile. the final column is dded to target.entropy
while(line!=''):
s=line.split()
s=[float(i) for i in s]
target.profile.append(s[0:21])
target.entropy.append(s[21])
line=handle.readline()
target.entropymean=np.mean(target.entropy) #Check what this does. I under stand the mean entropy is taken. Not sure on the rest.
target.potential=np.zeros(shape=(target.seqlen,target.seqlen))
target.mi=np.zeros(shape=(target.seqlen,target.seqlen))
target.minormal=np.zeros(shape=(target.seqlen,target.seqlen))
with open(pairwise_file,'rU') as handle: #Adds Frequency, mi and normalised mi
for line in handle:
s=line.split()
s=[float(i) for i in s]
target.potential[int(s[0])-1,int(s[1])-1]=s[2]
target.mi[int(s[0])-1,int(s[1])-1]=s[3]
target.minormal[int(s[0])-1,int(s[1])-1]=s[4]
target.psicov=np.zeros(shape=(target.seqlen,target.seqlen))
with open(psicov_file,'rU') as handle:
for line in handle:
s=line.split()
s=[float(i) for i in s]
target.psicov[int(s[0])-1,int(s[1])-1]=s[4]
target.evfold=np.zeros(shape=(target.seqlen,target.seqlen))
with open(evfold_file,'rU') as handle:
for line in handle:
s=line.split()
target.evfold[int(s[0])-1,int(s[2])-1]=float(s[5])
target.ccmpred=np.zeros(shape=(target.seqlen,target.seqlen))
with open(ccmpred_file,'rU') as handle:
indexx=0
for line in handle:
s=line.split()
for indexy,e in enumerate(s):
target.ccmpred[indexx,indexy]=e
indexx+=1
with open(seconds_file,'rU') as handle:
line=handle.readline()
line=handle.readline()
while(line!=''):
s=line.split()
if s:
target.ss_c.append(float(s[3]))
target.ss_h.append(float(s[4]))
target.ss_e.append(float(s[5]))
line=handle.readline()
with open(solvent_file,'rU') as handle:
line=handle.readline()
while(line!=''):
s=line.split()
target.solvent.append(float(s[3]))
line=handle.readline()
target.ss_cmean=np.mean(target.ss_c)
target.ss_hmean=np.mean(target.ss_h)
target.ss_emean=np.mean(target.ss_e)
target.solventmean=np.mean(target.solvent)
#print target.ss_c
for winpos in range(target.seqlen):
for winpos2 in range(winpos+5,target.seqlen):
activation=[0]*target.NUM_IN
for j in range(target.WINL,target.WINR+1):
if j+winpos>=0 and j+winpos<target.seqlen:
activation[(j-target.WINL)*target.IPERGRP]=target.ss_c[j+winpos]
activation[(j-target.WINL)*target.IPERGRP+1]=target.ss_h[j+winpos]
activation[(j-target.WINL)*target.IPERGRP+2]=target.ss_e[j+winpos]
activation[(j-target.WINL)*target.IPERGRP+3]=target.solvent[j+winpos]
activation[(j-target.WINL)*target.IPERGRP+4]=target.entropy[j+winpos]
else:
activation[(j-target.WINL)*target.IPERGRP+5]=1.0
for j in range(target.WINL,target.WINR+1):
if j+winpos2>=0 and j+winpos2<target.seqlen:
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP]=target.ss_c[j+winpos2]
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP+1]=target.ss_h[j+winpos2]
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP+2]=target.ss_e[j+winpos2]
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP+3]=target.solvent[j+winpos2]
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP+4]=target.entropy[j+winpos2]
else:
activation[(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.WINL)*target.IPERGRP+5]=1.0
midpos=int((winpos+winpos2)/2)
for j in range(target.CWINL,target.CWINR+1):
if j+midpos>=0 and j+midpos<target.seqlen:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP]=target.ss_c[j+midpos]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP+1]=target.ss_h[j+midpos]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP+2]=target.ss_e[j+midpos]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP+3]=target.solvent[j+midpos]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP+4]=target.entropy[j+midpos]
else:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(j-target.CWINL)*target.IPERGRP+5]=1.0
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP ]=target.mi[winpos,winpos2]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+1]=target.minormal[winpos,winpos2]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+2]=target.potential[winpos,winpos2]
count=0
for i in range(target.WINL,target.WINR+1):
for j in range(target.WINL,target.WINR+1):
if j+winpos2>=0 and j+winpos2<target.seqlen and i+winpos>=0 and i+winpos<target.seqlen:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+3+count]=target.ccmpred[winpos+i,winpos2+j]
count+=1
count1=0
for i in range(target.WINL,target.WINR+1):
for j in range(target.WINL,target.WINR+1):
if j+winpos2>=0 and j+winpos2<target.seqlen and i+winpos>=0 and i+winpos<target.seqlen:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+172+count1]=target.psicov[winpos+i,winpos2+j]
count1+=1
count2=0
for i in range(target.WINL,target.WINR+1):
for j in range(target.WINL,target.WINR+1):
if j+winpos2>=0 and j+winpos2<target.seqlen and i+winpos>=0 and i+winpos<target.seqlen:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+341+count2]=target.evfold[winpos+i,winpos2+j]
count2+=1
seqsep=winpos2-winpos
if seqsep<5:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+510]=1.0
elif seqsep<14:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+511]=1.0
elif seqsep<18:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+512]=1.0
elif seqsep<23:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+513]=1.0
elif seqsep<28:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+514]=1.0
elif seqsep<38:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+515]=1.0
elif seqsep<48:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+516]=1.0
else:
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+517]=1.0
for aa in range(21):
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+518+aa]=target.aacomposition[aa]
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+539]=target.ss_cmean
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+540]=target.ss_hmean
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+541]=target.ss_emean
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+542]=target.solventmean
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+543]=np.log(target.seqlen)
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+544]=np.log(target.nseq)
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+545]=np.log(target.effnseq)
activation[2*(target.WINR-target.WINL+1)*target.IPERGRP+(target.CWINR-target.CWINL+1)*target.IPERGRP+546]=target.entropymean
#produces 733 feature
outputhandle.write("%d\t%d\t"%(winpos,winpos2))
for e in activation:
outputhandle.write("%6.4f "%(e))
outputhandle.write("\n")
outputhandle.close()