|
30
|
1 #! /usr/bin/python
|
|
|
2 # -*- coding: utf8 -*-
|
|
|
3 """#Matrix Scan - developed by Jocelyn Brayet <jocelyn.brayet@curie.fr>
|
|
|
4 #Copyright (C) 2015 Institut Curie
|
|
|
5 #
|
|
|
6 #This program is free software: you can redistribute it and/or modify
|
|
|
7 #it under the terms of the GNU General Public License as published by
|
|
|
8 #the Free Software Foundation, either version 3 of the License, or
|
|
|
9 #(at your option) any later version.
|
|
|
10 #
|
|
|
11 #This program is distributed in the hope that it will be useful,
|
|
|
12 #but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
13 #MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
14 #GNU General Public License for more details.
|
|
|
15 #
|
|
|
16 #You should have received a copy of the GNU General Public License
|
|
|
17 #along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
18 #
|
|
|
19 ###########################################################'
|
|
|
20 #
|
|
|
21 #Client to download matrix scan results from RSAT server.
|
|
|
22 #
|
|
|
23 #
|
|
|
24 #usage: matrixScan_wrapper.py [-h] -sequence <SEQUENCE> -matrix <MATRIX>
|
|
|
25 # -n_treatment <N_TREATMENT> -markov <MARKOV>
|
|
|
26 # -matrix_format <MATRIX_FORMAT>
|
|
|
27 # [-organism <ORGANISM>] [-background <BACKGROUND>]
|
|
|
28 # [-origin <ORIGIN>] [-return_param <RETURN_PARAM>]
|
|
|
29 # [-uth <UTH>] -server <SERVEUR> -outGalaxy
|
|
|
30 # <OUT_GALAXY>
|
|
|
31 #
|
|
|
32 #optional arguments:
|
|
|
33 # -h, --help show this help message and exit
|
|
|
34 # -sequence <SEQUENCE>, --sequence <SEQUENCE>
|
|
|
35 # Sequence(s) to scan - all the formats supported in
|
|
|
36 # RSAT can be used as input (default: fasta).
|
|
|
37 # -matrix <MATRIX>, --matrix <MATRIX>
|
|
|
38 # Matrix/ces to scan with. The matrix format is
|
|
|
39 # specified with the option "matrix_format" (see below)
|
|
|
40 # Default format: tab.
|
|
|
41 # -n_treatment <N_TREATMENT>, --n_treatment <N_TREATMENT>
|
|
|
42 # Treatment of N characters. These characters are often
|
|
|
43 # used in DNA sequences to represent undefined or masked
|
|
|
44 # nucleotides (skip or score).
|
|
|
45 # -markov <MARKOV>, --markov <MARKOV>
|
|
|
46 # Order of the markov chain for the background model.
|
|
|
47 # -matrix_format <MATRIX_FORMAT>, --matrix_format <MATRIX_FORMAT>
|
|
|
48 # Supported fields: tab, cb, transfac, jaspar,
|
|
|
49 # consensus, gibbs, meme, assembly.
|
|
|
50 # -organism <ORGANISM>, --organism <ORGANISM>
|
|
|
51 # To use a precalculated background model from RSAT,
|
|
|
52 # choose the organism corresponding to the background
|
|
|
53 # model.
|
|
|
54 # -background <BACKGROUND>, --background <BACKGROUND>
|
|
|
55 # Type of sequences used as background model for
|
|
|
56 # estimating expected oligonucleotide frequencies.
|
|
|
57 # Supported: upstream, upstream-noorf
|
|
|
58 # -origin <ORIGIN>, --origin <ORIGIN>
|
|
|
59 # Define the origin for the calculation of positions.
|
|
|
60 # -return_param <RETURN_PARAM>, --return_param <RETURN_PARAM>
|
|
|
61 # List of fields to return.
|
|
|
62 # -uth <UTH>, --uth <UTH>
|
|
|
63 # Lower threshold on some parameter.
|
|
|
64 # -server <SERVEUR>, --server <SERVEUR>
|
|
|
65 # -outGalaxy <OUT_GALAXY>, --outGalaxy <OUT_GALAXY>
|
|
|
66 #
|
|
|
67 #Version 0.1 - 10/03/2015
|
|
|
68 #
|
|
|
69 ###########################################################"""
|
|
|
70 __author__ = 'Jocelyn Brayet'
|
|
|
71 matrixScanVersion = '0.1 - 10/03/2015'
|
|
|
72
|
|
|
73
|
|
|
74 ###########################################################'
|
|
|
75 ## Import
|
|
|
76
|
|
|
77 import argparse
|
|
|
78 import os
|
|
|
79 import urllib
|
|
|
80 from suds.client import Client
|
|
|
81 import platform
|
|
|
82
|
|
|
83 ###########################################################'
|
|
|
84
|
|
|
85 ################################ functions ############################################################
|
|
|
86 ## Define a function to make a service perform the desired request using provided arguments
|
|
|
87 def call_run_service(service, args):
|
|
|
88 """
|
|
|
89 Run job in RSAT server.
|
|
|
90 service -> RSAT web service
|
|
|
91 args -> web service request
|
|
|
92 """
|
|
|
93
|
|
|
94 result = rsat_service.matrix_scan(args)
|
|
|
95 return result
|
|
|
96
|
|
|
97 def testNone(argument):
|
|
|
98 """
|
|
|
99 Test if argument is None or not.
|
|
|
100 argument -> argument give by user
|
|
|
101 """
|
|
|
102
|
|
|
103 if not argument is None:
|
|
|
104 variable = argument[0]
|
|
|
105 else:
|
|
|
106 variable = ""
|
|
|
107 return variable
|
|
|
108
|
|
|
109 ###########################################################'
|
|
|
110
|
|
|
111 ###########################################################'
|
|
|
112 # server dictionary
|
|
|
113 serverDict = {
|
|
|
114
|
|
|
115 #http://protists.rsat.eu/
|
|
|
116 "fr_ens":"http://rsat01.biologie.ens.fr/rsa-tools/web_services/RSATWS.wsdl",
|
|
|
117 "fr_mrs":"http://rsat-tagc.univ-mrs.fr/rsat/web_services/RSATWS.wsdl",
|
|
|
118 "fr_ro":"http://rsat.sb-roscoff.fr/web_services/RSATWS.wsdl",
|
|
|
119 "fr_mrs_2":"http://pedagogix-tagc.univ-mrs.fr/rsat/web_services/RSATWS.wsdl",
|
|
|
120 "es":"http://floresta.eead.csic.es/rsat/web_services/RSATWS.wsdl",
|
|
|
121 "mx":"http://embnet.ccg.unam.mx/rsa-tools/web_services/RSATWS.wsdl"
|
|
|
122
|
|
|
123 }
|
|
|
124
|
|
|
125
|
|
|
126 if __name__ == '__main__':
|
|
|
127
|
|
|
128 parser = argparse.ArgumentParser(description='Client to download matrix-scan results from RSAT server.', epilog='Version '+matrixScanVersion)
|
|
|
129
|
|
|
130
|
|
|
131 ########### convert matrix arguments ####################
|
|
|
132
|
|
|
133 parser.add_argument('-sequence', '--sequence', metavar='<SEQUENCE>', type=argparse.FileType('r'), nargs=1, help='Sequence(s) to scan - all the formats supported in RSAT can be used as input (default: fasta).', required=True)
|
|
|
134 parser.add_argument('-matrix', '--matrix', metavar='<MATRIX>', type=argparse.FileType('r'), nargs=1, help='Matrix/ces to scan with. The matrix format is specified with the option "matrix_format" (see below) Default format: tab.', required=True)
|
|
|
135 parser.add_argument('-n_treatment', '--n_treatment', metavar='<N_TREATMENT>', type=str, nargs=1, help='Treatment of N characters. These characters are often used in DNA sequences to represent undefined or masked nucleotides (skip or score).', required=True)
|
|
|
136 parser.add_argument('-markov', '--markov', metavar='<MARKOV>', type=int, nargs=1, help='Order of the markov chain for the background model.', required=True)
|
|
|
137 parser.add_argument('-matrix_format', '--matrix_format', metavar='<MATRIX_FORMAT>', type=str, nargs=1, help='Supported fields: tab, cb, transfac, jaspar, consensus, gibbs, meme, assembly.', required=True)
|
|
|
138
|
|
|
139 parser.add_argument('-organism', '--organism', metavar='<ORGANISM>', type=str, nargs=1, help='To use a precalculated background model from RSAT, choose the organism corresponding to the background model.', required=False)
|
|
|
140 parser.add_argument('-background', '--background', metavar='<BACKGROUND>', type=str, nargs=1, help='Type of sequences used as background model for estimating expected oligonucleotide frequencies. Supported: upstream, upstream-noorf', required=False)
|
|
|
141 #parser.add_argument('-background_model', '--background_model', metavar='<BACKGROUND_MODEL>', type=str, nargs=1, help='Background model is a tab-delimited specification of oligonucleotide frequencies.', required=True)
|
|
|
142 parser.add_argument('-origin', '--origin', metavar='<ORIGIN>', type=str, nargs=1, help='Define the origin for the calculation of positions.', required=False)
|
|
|
143 parser.add_argument('-return_param', '--return_param', metavar='<RETURN_PARAM>', type=str, nargs=1, help='List of fields to return.', required=False)
|
|
|
144 parser.add_argument('-uth', '--uth', metavar='<UTH>', type=str, nargs=1, help='Lower threshold on some parameter.', required=False)
|
|
|
145
|
|
|
146 #parser.add_argument('-quick', '--quick', metavar='<QUICK>', type=str, help='Delegates scanning to the C program matrix-scan-quick.', required=True)
|
|
|
147
|
|
|
148
|
|
|
149 ########### galaxy arguments ##############################
|
|
|
150 parser.add_argument('-server', '--server', metavar='<SERVEUR>', type=str, nargs=1, required=True)
|
|
|
151 parser.add_argument('-outGalaxy', '--outGalaxy', metavar='<OUT_GALAXY>', type=str, nargs=1, required=True)
|
|
|
152 ###########################################################
|
|
|
153
|
|
|
154
|
|
|
155 args = parser.parse_args()
|
|
|
156
|
|
|
157
|
|
|
158
|
|
|
159
|
|
|
160 ###########################################################'
|
|
|
161
|
|
|
162 sequence_file = args.sequence[0].read()
|
|
|
163 matrix_file = args.matrix[0].read()
|
|
|
164 serverValue = testNone(args.server)
|
|
|
165 matrixFormatValue = testNone(args.matrix_format)
|
|
|
166 n_treatmentValue = testNone(args.n_treatment)
|
|
|
167 outGalaxyValue = testNone(args.outGalaxy)
|
|
|
168 markovValue = testNone(args.markov)
|
|
|
169 organismValue = testNone(args.organism)
|
|
|
170 backgroundValue = testNone(args.background)
|
|
|
171 originValue = testNone(args.origin)
|
|
|
172 returnValue = testNone(args.return_param)
|
|
|
173 uthValue = testNone(args.uth)
|
|
|
174
|
|
|
175 #backgroundModelValue = testNone(args.background_model)
|
|
|
176 #quickValue = testNone(args.quick)
|
|
|
177
|
|
|
178 ###########################################################'
|
|
|
179 ## Create the SOAP client to request the RSAT service
|
|
|
180
|
|
|
181
|
|
|
182 # Load Client class from suds
|
|
|
183
|
|
|
184 # Define URL for RSAT services
|
|
|
185 url = serverDict[serverValue]
|
|
|
186 # Create the client
|
|
|
187 client = Client(url)
|
|
|
188
|
|
|
189 # Need service interface to perform requests
|
|
|
190 rsat_service = client.service
|
|
|
191
|
|
|
192
|
|
|
193 #print client
|
|
|
194 #print(client.factory.create('matrix_scan'))
|
|
|
195
|
|
|
196 #request =
|
|
|
197 # (MatrixScanRequest){
|
|
|
198 # output = None
|
|
|
199 # sequence = None
|
|
|
200 # tmp_sequence_infile = None
|
|
|
201 # matrix = None
|
|
|
202 # tmp_matrix_infile = None
|
|
|
203 # sequence_format = None
|
|
|
204 # matrix_format = None
|
|
|
205 # quick = None
|
|
|
206 # n_treatment = None
|
|
|
207 # consensus_name = None
|
|
|
208 # pseudo = None
|
|
|
209 # equi_pseudo = None
|
|
|
210 # top_matrices = None
|
|
|
211 # background_model = None
|
|
|
212 # tmp_background_infile = None
|
|
|
213 # organism = None
|
|
|
214 # background = None
|
|
|
215 # background_input = None
|
|
|
216 # background_window = None
|
|
|
217 # markov = None
|
|
|
218 # background_pseudo = None
|
|
|
219 # return_fields = None
|
|
|
220 # sort_distrib = None
|
|
|
221 # lth[] = <empty>
|
|
|
222 # uth[] = <empty>
|
|
|
223 # str = None
|
|
|
224 # verbosity = None
|
|
|
225 # origin = None
|
|
|
226 # decimals = None
|
|
|
227 # crer_ids = None
|
|
|
228 # }
|
|
|
229 #}
|
|
|
230
|
|
|
231 # Define client header
|
|
|
232
|
|
|
233 userAgent = 'RSAT-Client/v%s (%s; Python %s; %s)' % (
|
|
|
234 matrixScanVersion,
|
|
|
235 os.path.basename( __file__ ),
|
|
|
236 platform.python_version(),
|
|
|
237 platform.system()
|
|
|
238 )
|
|
|
239
|
|
|
240 httpHeaders = {'User-agent': userAgent}
|
|
|
241 client.set_options(headers=httpHeaders)
|
|
|
242 client.set_options(timeout=300)
|
|
|
243
|
|
|
244
|
|
|
245 if not uthValue == "":
|
|
|
246 matrixScanRequest = {
|
|
|
247
|
|
|
248 #'output' : "ticket",
|
|
|
249 'sequence' : sequence_file,
|
|
|
250 'matrix' : matrix_file,
|
|
|
251 'markov' : markovValue,
|
|
|
252 'n_treatment' : n_treatmentValue,
|
|
|
253 'matrix_format' : matrixFormatValue,
|
|
|
254 'background' : backgroundValue,
|
|
|
255 'organism' : organismValue,
|
|
|
256
|
|
|
257 'origin' : originValue,
|
|
|
258 'return_fields' : returnValue,
|
|
|
259 'uth' : 'pval ' + uthValue,
|
|
|
260
|
|
|
261 #'background_window' : 12,
|
|
|
262 #'background_input' : " ",
|
|
|
263
|
|
|
264 'quick' : "-quick"
|
|
|
265
|
|
|
266 }
|
|
|
267
|
|
|
268 else:
|
|
|
269 matrixScanRequest = {
|
|
|
270
|
|
|
271 #'output' : "ticket",
|
|
|
272 'sequence' : sequence_file,
|
|
|
273 'matrix' : matrix_file,
|
|
|
274 'markov' : markovValue,
|
|
|
275 'n_treatment' : n_treatmentValue,
|
|
|
276 'matrix_format' : matrixFormatValue,
|
|
|
277 'background' : backgroundValue,
|
|
|
278 'organism' : organismValue,
|
|
|
279
|
|
|
280 'origin' : originValue,
|
|
|
281 'return_fields' : returnValue,
|
|
|
282 #'background_window' : 12,
|
|
|
283 #'background_input' : " ",
|
|
|
284
|
|
|
285 'quick' : "-quick"
|
|
|
286
|
|
|
287 }
|
|
|
288
|
|
|
289
|
|
|
290
|
|
|
291 #print matrixScanRequest
|
|
|
292
|
|
|
293 result = call_run_service(rsat_service, matrixScanRequest)
|
|
|
294
|
|
|
295 print url
|
|
|
296
|
|
|
297 print "###############################################"
|
|
|
298 print "Command performed on server"
|
|
|
299 print result.command
|
|
|
300 print "###############################################"
|
|
|
301 print "Result"
|
|
|
302 print result.server
|
|
|
303
|
|
|
304 nameFile = "matrix-scan_results.txt"
|
|
|
305
|
|
|
306 urlResult=result.server.replace("$RSAT/public_html/",url.replace("web_services/RSATWS.wsdl",""))
|
|
|
307
|
|
|
308
|
|
|
309 urllib.urlretrieve(urlResult, nameFile)
|
|
|
310
|
|
|
311
|
|
|
312 os.popen("cp "+nameFile+" "+outGalaxyValue)
|
|
|
313
|
|
|
314
|
