-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgeneticPooledOdds.py
330 lines (322 loc) · 18.1 KB
/
geneticPooledOdds.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
#!/usr/bin/python
#python geneticPooledOdds.py {ID} {File Name} {Age} {Gender} {Excel spreadsheet}
import ensemblRESTApi
import pandas as pd
import numpy as np
#import random
from datetime import datetime
import sys
import csv
from numerical.one_variable import secant
#excel sheet with info on first tab/sheet
excelSheetName = str(sys.argv[5])
fileName = str(sys.argv[2])
age = int(sys.argv[3])
gender = str(sys.argv[4])
idAtInvoke=sys.argv[1]
SNPdf = pd.read_excel(excelSheetName, sheetname = 'Master_SNP_List', parse_cols = 'A,C,E,M,N')
SNPdf['DBName'] = SNPdf['DBName'].astype(str)
SNPdf['DBName'] = SNPdf.DBName.str.strip(' ')
d = SNPdf.columns
d = ['rsid', 'Genotype', 'Disease', 'Reference', 'OR']
SNPdf.columns = d
SNPdf['rsid'] = SNPdf['rsid'].astype(str)
SNPdf.rsid = SNPdf.rsid.str.strip(' ').str.strip('\n')
SNPdf.Genotype = SNPdf.Genotype.astype(str)
SNPdf.Genotype = SNPdf.Genotype.str.strip(' ')
SNPdf['OR'] = SNPdf['OR'].astype(float)
series1 = pd.Series([0 for _ in range(len(SNPdf['Reference']))], index = SNPdf.index)
for i in range(len(series1)):
if SNPdf['Reference'].iloc[i] == "Yes":
series1.iloc[i] = 1
else:
series1.iloc[i] = 0
SNPdf.Reference = series1.astype(int)
prevalence = pd.read_excel(excelSheetName, sheetname = 'Sheet1', parse_cols = 'A,C,D,E,G')
prevalence.head()
prevalence
d = prevalence.columns
d = d.str.strip(' ')
prevalence.columns = d
prevalence.Disease = prevalence.Disease.astype(str)
prevalence.Disease = prevalence.Disease.str.strip(' ')
prevalence.Gender = prevalence.Gender.str.strip(' ')
prevalence.Prevalence = prevalence.Prevalence.astype(float)
prevalence.Age = prevalence.Age.astype(str)
prevalence = prevalence[prevalence['SampleSize'] != 'NaN']
series1 = pd.Series([0 for _ in range(len(prevalence['Gender']))], index = prevalence.index)
series2 = pd.Series([0 for _ in range(len(prevalence['Gender']))], index = prevalence.index)
for i in range(len(prevalence['Gender'])):
if prevalence['Gender'].iloc[i] == "Both":
series1.iloc[i] = 1
series2.iloc[i] = 1
if prevalence['Gender'].iloc[i] == "Male":
series1.iloc[i] = 1
series2.iloc[i] = 0
if prevalence['Gender'].iloc[i] == "Female":
series1.iloc[i] = 0
series2.iloc[i] = 1
prevalence["Male"] = series1.astype(int)
prevalence["Female"] = series2.astype(int)
series1 = pd.Series([0 for _ in range(len(prevalence['Gender']))], index = prevalence.index)
series2 = pd.Series([0 for _ in range(len(prevalence['Gender']))], index = prevalence.index)
for i in range(len(prevalence['Age'])):
def parseAge(a):
b = a
if '-' in a:
b = a.split('-')
if len(b) == 2:
series1.iloc[i] = b[0]
series2.iloc[i] = b[1]
else:
b = b.replace('+', '')
if b == 'nan':
series1.iloc[i] = 0
else:
series1.iloc[i] = int(b)
series2.iloc[i] = 200
parseAge(prevalence['Age'].iloc[i])
prevalence['MinAge'] = series1.astype(int)
prevalence['MaxAge'] = series2.astype(int)
# Reading source file
data = {}
with open(fileName) as file:
for line in file:
(key, val) = line.split(":")
key = key.replace('"', '')
key = key.replace(' ', '')
val = val.replace('"', '')
val = val.replace(' ', '')
val = val.replace('\n', '')
val = val.replace('\r', '')
# Getting rid of empty reads
if val != '__' and val != '--' and key != 'id':
data[key] = val
def PooledOddsForDisease(disease, PopulationProbability):
# print(disease)
subSNPdf = SNPdf[SNPdf['Disease'] == disease]
uniq = subSNPdf['rsid'].unique()
results = ensemblRESTApi.getGenotypeProbabilities(uniq)
SNPsPersonHas = list(set(data.keys()) & set(uniq))
DictionaryOfSNPAndGenotype = {}
for i in SNPsPersonHas:
DictionaryOfSNPAndGenotype[i] = data[i]
k = pd.DataFrame(list(data.items()), index = data.keys())
del k[0]
k.columns = [fileName]
def findConditionalProbabilities(popProb, OddsAndPGn):
if disease == "Cognitive_Function":
print(OddsAndPGn)
def findp0p1(d):
IsNegative = None
WasNegative = None
p0 = 0.0 #None
p1 = 1.0 #None
for i in range(len(d['Initial Candidates'])):
if d['Initial Candidates'].iloc[i] < 0:
IsNegative = True
if WasNegative == False:
p1 = d['Initial Candidates'].index[i]
elif p0 is None:
p0 = d['Initial Candidates'].index[i]
elif d['Initial Candidates'].iloc[i] > p0:
p0 = d['Initial Candidates'].index[i]
elif d['Initial Candidates'].iloc[i] > 0:
IsNegative = False
if WasNegative == True:
p1 = d['Initial Candidates'].index[i]
return p0, p1
elif p0 is None:
p0 = d['Initial Candidates'].index[i]
elif d['Initial Candidates'].iloc[i] < p0:
p0 = d['Initial Candidates'].index[i]
if IsNegative == True:
WasNegative = True
else:
WasNegative = False
return max(0, p0), min(p1, 1)
def computePDG_1(x):
result = x * OddsAndPGn['PGn'].iloc[0] - popProb
for i in range(1,len(OddsAndPGn)):
result += OddsAndPGn['OR'].iloc[i] * x / (OddsAndPGn['OR'].iloc[i] * x - x + 1) * OddsAndPGn['PGn'].iloc[i]
return result
df = pd.DataFrame([], index = np.linspace(0.0001, 1, 100), columns = ["Initial Candidates"])
for i in range(len(df)):
p_0 = df.index[i]
df["Initial Candidates"].iloc[i] = computePDG_1(p_0)
p0, p1 = findp0p1(df)
conditionalPDG_1 = secant(computePDG_1, p0, p1, 0.0001, 2000)
if conditionalPDG_1 > 0:
conditionalPDG_2 = OddsAndPGn['OR'].iloc[1] * conditionalPDG_1 / (OddsAndPGn['OR'].iloc[1] * conditionalPDG_1 - conditionalPDG_1 + 1)
conditionalPDG_3 = OddsAndPGn['OR'].iloc[2] * conditionalPDG_1 / (OddsAndPGn['OR'].iloc[2] * conditionalPDG_1 - conditionalPDG_1 + 1)
else:
conditionalPDG_1 = 0
conditionalPDG_2 = 0
conditionalPDG_3 = 0
return [conditionalPDG_1, conditionalPDG_2, conditionalPDG_3]
SNPDictionary = {}
#startTime = datetime.now()
for i in SNPsPersonHas:
def matchString(index, ts):
return (index == ts) | (index == ts[1] + ts[0])
def stripPipe(t):
temp = t.split('|')
return temp[0] + temp[1]
tempDF = subSNPdf[subSNPdf['rsid'] == i]
OddsAndPDns = pd.DataFrame(list(tempDF['OR']), index = tempDF['Genotype'], columns = ['OR'])
series = pd.Series([0 for _ in range(len(OddsAndPDns.index))], index = tempDF['Genotype'])
for k, j in enumerate(results[i]['population_genotypes']):
if 'ALL' in j['population']:
tempString = stripPipe(j['genotype'])
tempDF = SNPdf[(SNPdf['rsid'] == i) & (SNPdf['Genotype'] == tempString)]
if len(tempDF) == 0:
tempDF = SNPdf[SNPdf['rsid'] == i]
tempDF = tempDF[tempDF['Genotype'] == tempString[1] + tempString[0]]
if len(tempDF) == 1:
for o in range(len(OddsAndPDns)):
if matchString(OddsAndPDns.index[o], tempString):
series.iloc[o] = float(j['frequency'])
if disease == "Stroke_All":
print(series)
OddsAndPDns['OR'] = OddsAndPDns['OR'].astype(float)
OddsAndPDns['PGn'] = series.astype(float)
OddsAndPDns = OddsAndPDns.sort_values(['OR'], ascending = [True])
OddsAndPDns['OR'] = OddsAndPDns['OR'] / OddsAndPDns['OR'].iloc[0]
OddsAndPDns['Conditional'] = pd.Series(findConditionalProbabilities(PopulationProbability, OddsAndPDns), index = OddsAndPDns.index)
SNPDictionary[i] = OddsAndPDns
#print(datetime.now() - startTime)
denom = PopulationProbability / (1-PopulationProbability)
for i in SNPDictionary.keys():
product = 1
adjustedOR = []
for j in range(len(SNPDictionary[i]['Conditional'])):
tmp = SNPDictionary[i]['Conditional'].iloc[j]
#print(tmp / (1-tmp))
adjustedOR.append((tmp / (1-tmp)) / denom)
product *= ((tmp / (1-tmp)) / denom)
SNPDictionary[i]['AdjustedOR'] = pd.Series(adjustedOR, index = SNPDictionary[i].index)
pooledOdds = 1
for i in DictionaryOfSNPAndGenotype.keys():
if(len(SNPDictionary[i][SNPDictionary[i].index == DictionaryOfSNPAndGenotype[i]]['AdjustedOR']) == 1):
pooledOdds *= SNPDictionary[i][SNPDictionary[i].index == DictionaryOfSNPAndGenotype[i]]['AdjustedOR'].iloc[0]
return pooledOdds
diseasesList = prevalence['Disease'].unique()
outputDF = {}
for i, j in enumerate(diseasesList):
if j == "Allergic_Rhinitis":
outputDF[j] = PooledOddsForDisease(j, prevalence[prevalence['Disease'] == j]['Prevalence'].iloc[0])
elif j == "Alzheimer_Disease":
if age < 65:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == "65-69") & (prevalence['Gender'] == "Both")]['Prevalence'].iloc[0])
elif age > 65 & age < 90:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age) & (prevalence['Gender'] == 'Both')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (90 <= age) & (prevalence['Gender'] == 'Both')]['Prevalence'].iloc[0])
elif j == "Celiac_disease":
outputDF[j] = PooledOddsForDisease(j, prevalence[prevalence['Disease'] == j]['Prevalence'].iloc[0])
# Cognitive_Function is in Disease_Genetics.xlsx but contains no OR to use for calculation
#elif j == "Cognitive_Function":
# if age < 60:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] == 60)]['Prevalence'].iloc[0])
# else:
# k = prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0]
# print(j, k)
# outputDF[j] = PooledOddsForDisease(j, k)
elif j == "Coronary_Artery_Disease":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Type_2_Diabetes_Mellitus":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Gout":
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j)]['Prevalence'].iloc[0])
elif j == "Heart_Failure":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
# no odds ratios for the snps associated with Hyperhomocysteinemia
#elif j == "Hyperhomocysteinemia":
# if age < 67:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '67-74') & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
# else:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age) & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
# no odds ratios associated with Hyperlipidemia
#elif j == "Hyperlipidemia":
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j)]['Prevalence'].iloc[0])
#elif j == "Low_HDL":
# if age < 20:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '20-39') & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
# else:
# outputDF[j] = outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age) & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
#elif j == "Insomnia":
# outputDF[j] = PooledOddsForDisease(j, 0.299)
elif j == "Major_Depressive_Disorder":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-29')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Osteoarthritis":
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
elif j == "OsteoarthritisHip":
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender)]['Prevalence'].iloc[0])
elif j == "Osteoporosis":
if age < 50:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['Age'] == '50-59')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['MinAge'] <= age) & ( prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
#elif j == "Rheumatoid_arthritis":
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j)]['Prevalence'].iloc[0])
#Need to figure out what's going on with Stroke_All
#elif j == "Stroke_All":
# if age < 18:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
# else:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "StrokeLarge_Vessel":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
#elif j == "Ischemic_Stroke":
# if age < 18:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
# else:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Venous_thromboembolism":
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j)]['Prevalence'].iloc[0])
elif (j == "Polycystic_Ovary_Syndrome") & (gender == 'Female'):
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j)]['Prevalence'].iloc[0])
#elif j == "Hypertension":
# if age < 18:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
# else:
# outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Iron_Deficiency":
if gender == "Male":
if age < 12:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['Age'] == '12-15')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif gender == "Female":
if age < 12:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['Age'] == '12-15')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Gender'] == gender) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Migrane_Headache":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
elif j == "Obesity":
if age < 18:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['Age'] == '18-44')]['Prevalence'].iloc[0])
else:
outputDF[j] = PooledOddsForDisease(j, prevalence[(prevalence['Disease'] == j) & (prevalence['MinAge'] <= age) & (prevalence['MaxAge'] >= age)]['Prevalence'].iloc[0])
writer = csv.writer(open('{0}.csv'.format(str(idAtInvoke)), 'wb'))
for key, value in outputDF.items():
writer.writerow([key, value])