-
Notifications
You must be signed in to change notification settings - Fork 0
/
coFeatures.py
189 lines (163 loc) · 9.58 KB
/
coFeatures.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
########################################################################################
## BASIC LIBRARIES ##
import os
from os import path as p
import sys
import pandas as pd
import multiprocessing
import argpass
from itertools import product
import multiprocessing
from tqdm import tqdm
import yaml
## coFeatures SPESIFIC MODULES ##
from utils_coFeatures import *
from modules_coFeatures import *
########################################################################################
def read_inputs():
## create an argpass parser, read config file, snip off ".py" if on the end of file
parser = argpass.ArgumentParser()
parser.add_argument("--config")
args = parser.parse_args()
config=args.config
## Read config.yaml into a dictionary
with open(config,"r") as yamlFile:
config = yaml.safe_load(yamlFile)
pathInfo = config["pathInfo"]
cofactorInfo = config["cofactorInfo"]
optionsInfo = config["optionsInfo"]
return pathInfo,cofactorInfo, optionsInfo
########################################################################################
def main():
## LOAD USER INPUTS ##
pathInfo,cofactorInfo, optionsInfo = read_inputs()
# MAKE OUTPUT DIRECTORY ##
os.makedirs(pathInfo["outDir"],exist_ok=True)
# READ AMINO ACID TABLE INTO A DATAFRAME, GET LIST OF AMIO ACID NAMES ##
# GET LISTS OF PDB IDS AND PATHS
idList = getPdbList(pathInfo["inputDir"])
jobOrder = list(product(idList,optionsInfo["orbRange"],optionsInfo["cloudRange"]))
process_pdbs_multicore(jobOrder, pathInfo, cofactorInfo, optionsInfo)
# process_pdbs_singlecore(jobOrder, pathInfo, cofactorInfo, optionsInfo)
merge_temporary_csvs(outDir = pathInfo["outDir"],
orbRange = optionsInfo["orbRange"],
cloudRange = optionsInfo["cloudRange"])
print("\nAll features have been generated and saved!")
########################################################################################
def process_pdbs_singlecore(jobOrder, pathInfo, cofactorInfo, optionsInfo):
aminoAcidNames, aminoAcidProperties = initialiseAminoAcidInformation(pathInfo["aminoAcidTable"])
for jobDetails in jobOrder:
process_pdbs_worker(jobDetails, pathInfo, cofactorInfo, optionsInfo, aminoAcidNames, aminoAcidProperties)
########################################################################################
def process_pdbs_multicore(jobOrder, pathInfo, cofactorInfo, optionsInfo):
aminoAcidNames, aminoAcidProperties = initialiseAminoAcidInformation(pathInfo["aminoAcidTable"])
num_processes = multiprocessing.cpu_count()
with multiprocessing.Pool(processes=num_processes) as pool:
pool.starmap(process_pdbs_worker,
tqdm( [(jobDetails, pathInfo, cofactorInfo, optionsInfo, aminoAcidNames, aminoAcidProperties) for jobDetails in jobOrder],
total = len(jobOrder)))
########################################################################################
def process_pdbs_worker(jobDetails, pathInfo, cofactorInfo, optionsInfo, aminoAcidNames, aminoAcidProperties):
## UNPACK pathInfo
pdbDir = pathInfo["inputDir"]
outDir = pathInfo["outDir"]
## UNPACK JOB DETAILS INTO VARIABLES ##
pdbId, orbValue, cloudValue = jobDetails
pdbFile = p.join(pdbDir,f"{pdbId}.pdb")
## INITIALSE TEMPORARY OUTPUT FILE, SKIP IF IT ALREADY EXISTS ##
outputCsv=p.join(outDir, f"{pdbId}_{str(orbValue)}_{str(cloudValue)}.tmp")
if p.isfile(outputCsv):
return
# INITIALISE LIST TO STORE ALL FEATURE DATAFRAMES ##
pdbDf=pdb2df(pdbFile)
cofactorCount, cofactorNames = find_cofactor(pdbDf=pdbDf,cofactorInfo=cofactorInfo)
## SKIP ZERO COFACTORS OR MORE THAN ONE TYPE PRESENT ##
if cofactorCount == 0:
return
elif len(cofactorNames) > 1:
return
cofactorName = cofactorNames[0]
## GET ORB, CLOUD, AND PROTEIN REGION DATAFRAMES ##
orbDf = gen_orb_region(orbAtoms=cofactorInfo[cofactorName]["orbAtoms"],
cofactorName=cofactorName,
pdbDf=pdbDf,
orbValue=orbValue)
cloudDf = gen_cloud_region(cloudAtoms=cofactorInfo[cofactorName]["cloudAtoms"],
cofactorName=cofactorName,
pdbDf=pdbDf,
cloudValue=cloudValue)
proteinDf = gen_protein_region(pdbDf=pdbDf,
cofactorName=cofactorName)
## COUNT ELEMENTS IN REGIONS ##
orbElementCountDf = element_count_in_region(regionDf=orbDf,
regionName="orb",
proteinName=pdbId)
cloudElementCountDf = element_count_in_region(regionDf=cloudDf,
regionName="cloud",
proteinName=pdbId)
proteinElementCountDf = element_count_in_region(regionDf=proteinDf,
regionName="protein",
proteinName=pdbId)
## COUNT AMINO ACIDS IN REGIONS ##
orbAACountDf = amino_acid_count_in_region(regionDf=orbDf,
regionName="orb",
proteinName=pdbId,
aminoAcidNames=aminoAcidNames)
cloudAACountDf = amino_acid_count_in_region(regionDf=cloudDf,
regionName="cloud",
proteinName=pdbId,
aminoAcidNames=aminoAcidNames)
proteinAACountDf = amino_acid_count_in_region(regionDf=proteinDf,
regionName="protein",
proteinName=pdbId,
aminoAcidNames=aminoAcidNames)
## AMINO ACID PROPERTIES FOR REGIONS ##
orbPropertiesDf = calculate_amino_acid_properties_in_region(aaCountDf=orbAACountDf,
aminoAcidProperties=aminoAcidProperties,
aminoAcidNames=aminoAcidNames,
proteinName=pdbId,
regionName="orb")
cloudPropertiesDf = calculate_amino_acid_properties_in_region(aaCountDf=cloudAACountDf,
aminoAcidProperties=aminoAcidProperties,
aminoAcidNames=aminoAcidNames,
proteinName=pdbId,
regionName="cloud")
proteinPropertiesDf = calculate_amino_acid_properties_in_region(aaCountDf=proteinAACountDf,
aminoAcidProperties=aminoAcidProperties,
aminoAcidNames=aminoAcidNames,
proteinName=pdbId,
regionName="protein")
## EXTRACT COORDINATES OF USER-DEFINED KEY ATOMS ##
keyAtomCoordsList = get_key_atom_coords(pdbDf = pdbDf,
keyAtoms = cofactorInfo[cofactorName]["keyAtoms"],
cofactorName = cofactorName)
## NEAREST AMINO ACIDS TO KEY ATOMS ##
keyAtomsFeaturesDf = nearest_n_residues_to_key_atom(keyAtomCoordsList=keyAtomCoordsList,
pdbDf=pdbDf,
aminoAcidNames=aminoAcidNames,
aminoAcidProperties=aminoAcidProperties,
proteinName=pdbId,
cofactorName=cofactorName,
nNearestList=[1,3])
# combine features dataframes and
featuresToConcat = [orbElementCountDf, cloudElementCountDf, proteinElementCountDf,
orbAACountDf, cloudAACountDf, proteinAACountDf,
orbPropertiesDf, cloudPropertiesDf, proteinPropertiesDf,
keyAtomsFeaturesDf]
featuresDf=pd.concat(featuresToConcat,axis=1)
if optionsInfo["genAminoAcidCategories"]:
featuresDf = make_amino_acid_category_counts(dataDf = featuresDf,
optionsInfo = optionsInfo)
if optionsInfo["normaliseCounts"]:
featuresDf = normalise_counts_by_size(dataDf = featuresDf,
aminoAcidNames = aminoAcidNames,
optionsInfo= optionsInfo)
## normalise by cofactor count
featuresDf = featuresDf / cofactorCount
## write to file
tmpSaveFile = p.join(outDir,f"{pdbId}_{str(orbValue)}_{str(cloudValue)}.csv")
featuresDf.to_csv(tmpSaveFile,index=True, sep=",")
return
########################################################################################
if __name__ == "__main__":
main()