-
Notifications
You must be signed in to change notification settings - Fork 0
/
EPP.py
363 lines (332 loc) · 16.6 KB
/
EPP.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
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
import importlib
# Install required packages
packages = ['pandas', 'numpy', 'scipy', 'scikit-posthocs', 'peptides']
for p in packages:
try:
importlib.import_module(p)
print(f"{p} is already installed.")
except ImportError:
print(f"{p} is not installed - now installing...")
try:
import subprocess
subprocess.check_call(['pip', 'install', p])
print(f"{p} has been successfully installed.")
except Exception as e:
print(f"Failed to install {p}: {e}")
# Import and abbreviate the packages
import pandas as pd
import numpy as np
import re
from scipy.stats import shapiro, f_oneway, kruskal, levene, ttest_ind, mannwhitneyu
from scikit_posthocs import posthoc_dunn, posthoc_tukey_hsd
from peptides import Peptide
# Writes the core allele dictionary of control and risk HLA alleles
def allele_types(control_alleles, risk_alleles):
allele_dictionary = {}
for c in control_alleles:
allele_dictionary[c] = 'Control'
for r in risk_alleles:
allele_dictionary[r] = 'Risk'
return allele_dictionary
# Formats the IEDB output and creates MixMHC2pred input
def iedb_format(iedb_csv, plist, allele_dictionary):
data = pd.read_csv(iedb_csv)
data.rename(columns={'rank': 'adjusted_rank'}, inplace=True)
filter = (data['adjusted_rank'] <= 20)
fd = data[filter]
fd = fd.rename(columns={'seq_num': 'source_protein'})
values = fd['source_protein'].unique()
sorted_values = sorted(values)
fd['source_protein'].replace(sorted_values, plist, inplace=True)
fd['allele_type'] = fd['allele'].map(allele_dictionary)
variable_column = fd.pop('allele_type')
fd.insert(1, 'allele_type', variable_column)
list = fd.peptide.unique()
np.asarray(list)
np.savetxt('HLAII_peptide_output.txt', list, fmt='%s')
return fd
# Formats the MixMHC2pred output and merges the output
def merge_data(iedb_csv, mix_csv, allele_dictionary):
with open(mix_csv, "r") as mix_input:
lines = mix_input.readlines()[19:]
with open('mix_csv_formatted', "w") as output_file:
output_file.writelines(lines)
data2 = pd.read_csv('mix_csv_formatted', delimiter="\t")
data2['BestAllele_type'] = data2['BestAllele'].map(allele_dictionary)
best_type_column = data2.pop('BestAllele_type')
data2.insert(3, 'BestAllele_type', best_type_column)
data2 = data2.rename(columns={'Peptide': 'peptide'})
merged_data = pd.merge(iedb_csv, data2, on='peptide')
merged_data.sort_values('%Rank_best', ascending=True, inplace=True)
protein_list = merged_data.source_protein.unique()
DfD = {elem: pd.DataFrame() for elem in protein_list}
for key in DfD.keys():
DfD[key] = merged_data[:][merged_data.source_protein == key]
DfD[key] = DfD[key].sort_values(['start', 'end', 'adjusted_rank'])
return merged_data, DfD
def pdif(DfD, allele_dictionary):
def hla_recognise(ad):
# Create dictionaries for storing the different HLA-II alleles
dp_strings = {}
dm_strings = {}
do_strings = {}
dq_strings = {}
dr_strings = {}
# Create regex expressions for the subtypes
dp_pattern = re.compile(r'DP')
dm_pattern = re.compile(r'DM')
do_pattern = re.compile(r'DO')
dq_pattern = re.compile(r'DQ')
dr_pattern = re.compile(r'DR')
# Execute pattern recognition and add to dictionaries
for key, value in ad.items():
if re.search(dp_pattern, key):
dp_strings[key] = value
elif re.search(dm_pattern, key):
dm_strings[key] = value
elif re.search(do_pattern, key):
do_strings[key] = value
elif re.search(dq_pattern, key):
dq_strings[key] = value
elif re.search(dr_pattern, key):
dr_strings[key] = value
return dp_strings, dm_strings, do_strings, dq_strings, dr_strings
dp, dm, do, dq, dr = hla_recognise(allele_dictionary)
subtype_dict = {'dp': dp, 'dm': dm, 'do': do, 'dq': dq, 'dr': dr}
test_alleles = {subtype: alleles for subtype, alleles in subtype_dict.items() if len(alleles) >= 2}
pdif_output = pd.DataFrame(columns=['protein'])
pdif_phs = {} # Dictionary for post-hocs
for subtype, alleles in test_alleles.items():
allele_list = list(alleles.keys())
if f'{subtype}_test' not in pdif_output.columns:
pdif_output[f'{subtype}_test'] = None
pdif_output[f'{subtype}_p'] = None
pdif_output[f'{subtype}_sig'] = None
for key in DfD:
data = DfD[key].loc[DfD[key]['allele'].isin(allele_list), ['adjusted_rank', 'allele']]
if pdif_output[pdif_output['protein'] == key].empty:
pdif_output = pdif_output.append({'protein': key}, ignore_index=True)
temp = []
violated = False
for allele in allele_list:
column_data = data[data['allele'] == allele]['adjusted_rank']
if len(column_data) >= 3: # Test how many entries there are in the dataset
temp.append(column_data)
stat, p = shapiro(column_data)
if p >= 0.05:
violated = True
else:
temp.append(column_data)
violated = True
if len(allele_list) > 2:
if not violated:
stat, levene_p = levene(*temp) # Conduct levenes test if data is normal
if levene_p < 0.05:
violated = True
if not violated: # Conduct ANOVA if normal & homogenous variances
stat, p = f_oneway(*temp)
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_test'] = 'ANOVA'
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_p'] = p
if p <= 0.05:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'significant'
posthoc = posthoc_tukey_hsd(temp)
posthoc = posthoc.rename(columns=dict(zip(posthoc.columns, allele_list)),
index=dict(zip(posthoc.index, allele_list)))
posthoc['total'] = posthoc.apply(lambda row: sum(row <= 0.05), axis=1)
pdif_phs[key] = posthoc
else:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'ns'
else: # Kruskal-Wallis if criteria violated
stat, p = kruskal(*temp)
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_test'] = 'Kruskal'
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_p'] = p
if p <= 0.05:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'significant'
posthoc = posthoc_dunn(temp, p_adjust='bonferroni')
posthoc = posthoc.rename(columns=dict(zip(posthoc.columns, allele_list)),
index=dict(zip(posthoc.index, allele_list)))
pdif_phs[key] = posthoc
else:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'ns'
else:
if not violated:
stat, levene_p = levene(*temp) # Conduct levenes test if data is normal
if levene_p < 0.05:
violated = True
if not violated: # Conduct T-test if normality satisfied
stat, p = ttest_ind(*temp) # T-test if criteria satisfied
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_test'] = 't_test'
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_p'] = p
if p <= 0.05:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'significant'
else:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'ns'
else:
stat, p = mannwhitneyu(*temp) # MannwhitneyU if criteria violated
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_test'] = 'mannwhitneyu'
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_p'] = p
if p <= 0.05:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'significant'
else:
pdif_output.loc[pdif_output['protein'] == key, f'{subtype}_sig'] = 'ns'
return pdif_output, pdif_phs
def adif(merged_data):
allele_list = merged_data.allele.unique()
AD = {elem: pd.DataFrame() for elem in allele_list}
for key in AD.keys():
AD[key] = merged_data[:][merged_data.allele == key]
temp = []
adif_output = pd.DataFrame(columns=['Allele', 'test', 'p_value', 'sig'])
adif_phs = {}
for key in AD:
data = AD[key].loc[:, ['adjusted_rank', 'source_protein']]
adif_output = adif_output.append({'Allele': key}, ignore_index=True)
proteins = AD[key].source_protein.unique()
violated = False
for protein in proteins:
column_data = data[data['source_protein'] == protein]['adjusted_rank']
temp.append(column_data)
if len(column_data) < 3:
violated = True
else:
stat, p = shapiro(column_data)
if p >= 0.05:
violated = True
if not violated:
stat, levene_p = levene(*temp)
if levene_p < 0.05:
violated = True
if not violated and len(proteins) > 2:
stat, p = f_oneway(*temp) # ANOVA if criteria satisfied
adif_output.loc[adif_output['Allele'] == key, 'test'] = 'ANOVA'
if p <= 0.05:
posthoc = posthoc_tukey_hsd(temp)
posthoc = posthoc.rename(columns=dict(zip(posthoc.columns, proteins)),
index=dict(zip(posthoc.index, proteins)))
posthoc['total'] = posthoc.apply(lambda row: sum(row <= 0.05), axis=1)
adif_phs[key] = posthoc
if not violated and len(proteins) <= 2:
stat, p = ttest_ind(*temp) # T-test if criteria satisfied
adif_output.loc[adif_output['Allele'] == key, 'test'] = 't_test'
if violated and len(proteins) > 2:
stat, p = kruskal(*temp) # Kruskal-Wallis if criteria violated
adif_output.loc[adif_output['Allele'] == key, 'test'] = 'Kruskal'
if p <= 0.05:
posthoc = posthoc_dunn(temp, p_adjust='fdr_bh')
posthoc = posthoc.rename(columns=dict(zip(posthoc.columns, proteins)),
index=dict(zip(posthoc.index, proteins)))
posthoc['total'] = posthoc.apply(lambda row: sum(row <= 0.05), axis=1)
adif_phs[key] = posthoc
if violated and len(proteins) <= 2:
stat, p = mannwhitneyu(*temp)
adif_output.loc[adif_output['Allele'] == key, 'test'] = 'mannwhitneyu'
temp = []
adif_output.loc[adif_output['Allele'] == key, 'p_value'] = p
if p <= 0.05:
adif_output.loc[adif_output['Allele'] == key, 'sig'] = 'sig'
else:
adif_output.loc[adif_output['Allele'] == key, 'sig'] = 'ns'
adif_output.sort_values('p_value', ascending=True, inplace=True)
for protein in proteins:
temp_mean = data[data['source_protein'] == protein]['adjusted_rank'].mean()
protein_col = protein + '_mean'
adif_output.loc[adif_output['Allele'] == key, protein_col] = temp_mean
return adif_output, adif_phs
def identify_regions(merged_data, target_proteins, allele_dictionary, length=15, abv_proportion=0.7, peptide_pH=7.4):
targets = [target_proteins]
length -= 1
regions = {elem: pd.DataFrame() for elem in targets}
peptide = {elem: pd.DataFrame() for elem in targets}
class hla_allele:
def __init__(self, a, b, c):
self.subtype = a
self.name = b
self.type = c
def hla_recognise(ad):
allele_list = [] # Initialize as a list
# Create regex expressions for the subtypes
dp_pattern = re.compile(r'DP')
dm_pattern = re.compile(r'DM')
do_pattern = re.compile(r'DO')
dq_pattern = re.compile(r'DQ')
dr_pattern = re.compile(r'DR')
# Execute pattern recognition and add to list
for key, value in ad.items():
if re.search(dp_pattern, key):
subtype = 'dp'
elif re.search(dm_pattern, key):
subtype = 'dm'
elif re.search(do_pattern, key):
subtype = 'do'
elif re.search(dq_pattern, key):
subtype = 'dq'
elif re.search(dr_pattern, key):
subtype = 'dr'
temp = hla_allele(subtype, key, value)
allele_list.append(temp)
return allele_list
allele_list = hla_recognise(allele_dictionary)
for alleles in allele_list:
for target in targets:
# 1 Identify protein regions
region_output = pd.DataFrame(columns=['RegionID', 'RegionStart', 'RegionEnd', 'allele', 'allele_type'])
if alleles.type == 'Risk':
indicator = 'r'
else:
indicator = 'c'
data = merged_data[merged_data['allele'] == alleles.name]
RegionID = 0
RegionStart = data['start'].min()
RegionEnd = RegionStart + length
while True:
subset = data.loc[(data['start'] >= RegionStart) & (data['start'] <= RegionEnd), 'start']
if not subset.empty:
max_start = max(subset)
new_RegionEnd = max_start + length
if new_RegionEnd <= RegionEnd:
RegionID += 1
region_output = region_output.append(
{'RegionID': f'{target}_{indicator}{RegionID}', 'RegionStart': RegionStart,
'RegionEnd': RegionEnd, 'allele': alleles.name, 'allele_type': alleles.type},
ignore_index=True)
RegionStart = data.loc[data['start'] > RegionEnd, 'start'].min()
RegionEnd = RegionStart + length
continue
else:
RegionEnd = new_RegionEnd
else:
break
target_dict = regions.setdefault(target, {})
subtype_list = target_dict.setdefault(alleles.subtype, [])
subtype_list.append(region_output)
return regions
"""
# 2 Identify candidate peptides using protein ABVs
peptide_output = pd.DataFrame()
db_peptides = DfDRB[target]
for index, row in db_peptides.iterrows():
if row['allele_type'] == 'Risk':
risk_peptide = row['peptide']
control_row = db_peptides[
(db_peptides['allele_type'] == 'Control') & (db_peptides['peptide'] == risk_peptide)]
if control_row.empty:
peptide_output = peptide_output.append(row)
else:
risk_adjusted_rank = row['adjusted_rank']
control_adjusted_rank = control_row['adjusted_rank'].iloc[0]
if risk_adjusted_rank <= abv_proportion * control_adjusted_rank:
peptide_output = peptide_output.append(row)
file_name = 'DRB_' + str(target) + '_peptides.txt'
peptide_output['peptide'].to_csv(f'Peptides/DRB/{file_name}', index=False, header=True)
# 3 Predict the peptide properties using peptides.py
pep_df = pd.DataFrame(columns=['peptide', 'charge', 'hydrophobicity', 'instability',
'isoelectric_point', 'molecular_weight', 'mz'])
for index, row in peptide_output.iterrows():
peptide = Peptide(row['peptide'])
pep_df.loc[index] = [row['peptide'], peptide.charge(pH=peptide_pH, pKscale="EMBOSS"),
peptide.hydrophobicity(scale="Aboderin"), round(peptide.instability_index(), 2),
peptide.isoelectric_point(pKscale="EMBOSS"), peptide.molecular_weight(), peptide.mz()]
peptide[target] = pep_df
regions[target] = peptide_output
regions[target].sort_values('adjusted_rank', ascending=True, inplace=True)
"""