Mercurial > repos > ieguinoa > data_manager_salmon_index_builder

comparison data_manager/data_manager_fetch_gff.py @ 0:6cd60ba8a842 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author ieguinoa
date Tue, 14 Aug 2018 11:14:52 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:6cd60ba8a842
1 #!/usr/bin/env python
2 #Dan Blankenberg
3
4 import sys
5 import os
6 import tempfile
7 import shutil
8 import optparse
9 from ftplib import FTP
10 import tarfile
11 import zipfile
12 import gzip
13 import bz2
14 try:
15 # For Python 3.0 and later
16 from urllib.request import urlopen
17 from io import BytesIO as StringIO
18 from io import UnsupportedOperation
19 except ImportError:
20 # Fall back to Python 2's urllib2
21 from urllib2 import urlopen
22 from StringIO import StringIO
23 UnsupportedOperation = AttributeError
24 from json import loads, dumps
25
26
27 CHUNK_SIZE = 2**20 # 1mb
28
29 DATA_TABLE_NAME = 'all_gff'
30
31 def cleanup_before_exit( tmp_dir ):
32 if tmp_dir and os.path.exists( tmp_dir ):
33 shutil.rmtree( tmp_dir )
34
35
36 def stop_err(msg):
37 sys.stderr.write(msg)
38 sys.exit(1)
39
40
41 def get_dbkey_dbname_id_name( params, dbkey_description=None ):
42 # dbkey = params['param_dict']['dbkey_source']['dbkey']
43 #TODO: ensure sequence_id is unique and does not already appear in location file
44 sequence_id = params['param_dict']['sequence_id']
45 if not sequence_id:
46 sequence_id = dbkey #uuid.uuid4() generate and use an uuid instead?
47
48 # if params['param_dict']['dbkey_source']['dbkey_source_selector'] == 'new':
49 # dbkey_name = params['param_dict']['dbkey_source']['dbkey_name']
50 # if not dbkey_name:
51 # dbkey_name = dbkey
52 # else:
53 # dbkey_name = None
54 dbkey = params['param_dict']['dbkey']
55 dbkey_name = dbkey_description
56 sequence_name = params['param_dict']['sequence_name']
57 if not sequence_name:
58 sequence_name = dbkey_description
59 if not sequence_name:
60 sequence_name = dbkey
61 return dbkey, dbkey_name, sequence_id, sequence_name
62
63
64 def _get_files_in_ftp_path( ftp, path ):
65 path_contents = []
66 ftp.retrlines( 'MLSD %s' % ( path ), path_contents.append )
67 return [ line.split( ';' )[ -1 ].lstrip() for line in path_contents ]
68
69
70 def _get_stream_readers_for_tar( fh, tmp_dir ):
71 fasta_tar = tarfile.open( fileobj=fh, mode='r:*' )
72 return [x for x in [fasta_tar.extractfile(member) for member in fasta_tar.getmembers()] if x]
73
74
75 def _get_stream_readers_for_zip( fh, tmp_dir ):
76 """
77 Unpacks all archived files in a zip file.
78 Individual files will be concatenated (in _stream_fasta_to_file)
79 """
80 fasta_zip = zipfile.ZipFile( fh, 'r' )
81 rval = []
82 for member in fasta_zip.namelist():
83 fasta_zip.extract( member, tmp_dir )
84 rval.append( open( os.path.join( tmp_dir, member ), 'rb' ) )
85 return rval
86
87
88 def _get_stream_readers_for_gzip( fh, tmp_dir ):
89 return [ gzip.GzipFile( fileobj=fh, mode='rb') ]
90
91
92 def _get_stream_readers_for_bz2( fh, tmp_dir ):
93 return [ bz2.BZ2File( fh.name, 'rb') ]
94
95
96 def sort_fasta( fasta_filename, sort_method, params ):
97 if sort_method is None:
98 return
99 assert sort_method in SORTING_METHODS, ValueError( "%s is not a valid sorting option." % sort_method )
100 return SORTING_METHODS[ sort_method ]( fasta_filename, params )
101
102
103 def _move_and_index_fasta_for_sorting( fasta_filename ):
104 unsorted_filename = tempfile.NamedTemporaryFile().name
105 shutil.move( fasta_filename, unsorted_filename )
106 fasta_offsets = {}
107 unsorted_fh = open( unsorted_filename )
108 while True:
109 offset = unsorted_fh.tell()
110 line = unsorted_fh.readline()
111 if not line:
112 break
113 if line.startswith( ">" ):
114 line = line.split( None, 1 )[0][1:]
115 fasta_offsets[ line ] = offset
116 unsorted_fh.close()
117 current_order = map( lambda x: x[1], sorted( map( lambda x: ( x[1], x[0] ), fasta_offsets.items() ) ) )
118 return ( unsorted_filename, fasta_offsets, current_order )
119
120
121 def _write_sorted_fasta( sorted_names, fasta_offsets, sorted_fasta_filename, unsorted_fasta_filename ):
122 unsorted_fh = open( unsorted_fasta_filename )
123 sorted_fh = open( sorted_fasta_filename, 'wb+' )
124
125 for name in sorted_names:
126 offset = fasta_offsets[ name ]
127 unsorted_fh.seek( offset )
128 sorted_fh.write( unsorted_fh.readline() )
129 while True:
130 line = unsorted_fh.readline()
131 if not line or line.startswith( ">" ):
132 break
133 sorted_fh.write( line )
134 unsorted_fh.close()
135 sorted_fh.close()
136
137
138 def _sort_fasta_as_is( fasta_filename, params ):
139 return
140
141 def _sort_fasta_lexicographical( fasta_filename, params ):
142 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
143 sorted_names = sorted( fasta_offsets.keys() )
144 if sorted_names == current_order:
145 shutil.move( unsorted_filename, fasta_filename )
146 else:
147 _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
148
149
150 def _sort_fasta_gatk( fasta_filename, params ):
151 #This method was added by reviewer request.
152 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
153 sorted_names = map( str, range( 1, 23 ) ) + [ 'X', 'Y' ]
154 #detect if we have chrN, or just N
155 has_chr = False
156 for chrom in sorted_names:
157 if "chr%s" % chrom in current_order:
158 has_chr = True
159 break
160
161 if has_chr:
162 sorted_names = map( lambda x: "chr%s" % x, sorted_names)
163 sorted_names.insert( 0, "chrM" )
164 else:
165 sorted_names.insert( 0, "MT" )
166 sorted_names.extend( map( lambda x: "%s_random" % x, sorted_names ) )
167
168 existing_sorted_names = []
169 for name in sorted_names:
170 if name in current_order:
171 existing_sorted_names.append( name )
172 for name in current_order:
173 #TODO: confirm that non-canonical names do not need to be sorted specially
174 if name not in existing_sorted_names:
175 existing_sorted_names.append( name )
176
177 if existing_sorted_names == current_order:
178 shutil.move( unsorted_filename, fasta_filename )
179 else:
180 _write_sorted_fasta( existing_sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
181
182
183 def _sort_fasta_custom( fasta_filename, params ):
184 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
185 sorted_names = []
186 for id_repeat in params['param_dict']['sorting']['sequence_identifiers']:
187 sorted_names.append( id_repeat[ 'identifier' ] )
188 handle_not_listed = params['param_dict']['sorting']['handle_not_listed_selector']
189 if handle_not_listed.startswith( 'keep' ):
190 add_list = []
191 for name in current_order:
192 if name not in sorted_names:
193 add_list.append( name )
194 if add_list:
195 if handle_not_listed == 'keep_append':
196 sorted_names.extend( add_list )
197 else:
198 add_list.extend( sorted_names )
199 sorted_names = add_list
200 if sorted_names == current_order:
201 shutil.move( unsorted_filename, fasta_filename )
202 else:
203 _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
204
205
206 def _download_file(start, fh):
207 tmp = tempfile.NamedTemporaryFile()
208 tmp.write(start)
209 tmp.write(fh.read())
210 tmp.flush()
211 tmp.seek(0)
212 return tmp
213
214
215 def get_stream_reader(fh, tmp_dir):
216 """
217 Check if file is compressed and return correct stream reader.
218 If file has to be downloaded, do it now.
219 """
220 magic_dict = {
221 b"\x1f\x8b\x08": _get_stream_readers_for_gzip,
222 b"\x42\x5a\x68": _get_stream_readers_for_bz2,
223 b"\x50\x4b\x03\x04": _get_stream_readers_for_zip,
224 }
225 start_of_file = fh.read(CHUNK_SIZE)
226 try:
227 fh.seek(0)
228 except UnsupportedOperation: # This is if fh has been created by urlopen
229 fh = _download_file(start_of_file, fh)
230 for k,v in magic_dict.items():
231 if start_of_file.startswith(k):
232 return v(fh, tmp_dir)
233 try: # Check if file is tar file
234 if tarfile.open(fileobj=StringIO(start_of_file)):
235 return _get_stream_readers_for_tar(fh, tmp_dir)
236 except tarfile.ReadError:
237 pass
238 return fh
239
240
241 def _get_ucsc_download_address(params, dbkey):
242 """
243 Check if we can find the correct file for the supplied dbkey on UCSC's FTP server
244 """
245 UCSC_FTP_SERVER = 'hgdownload.cse.ucsc.edu'
246 UCSC_DOWNLOAD_PATH = '/goldenPath/%s/bigZips/'
247 COMPRESSED_EXTENSIONS = ['.tar.gz', '.tgz', '.tar.bz2', '.zip', '.fa.gz', '.fa.bz2']
248
249 email = params['param_dict']['__user_email__']
250 if not email:
251 email = 'anonymous@example.com'
252
253 ucsc_dbkey = params['param_dict']['reference_source']['requested_dbkey'] or dbkey
254 UCSC_CHROM_FA_FILENAMES = ['%s.chromFa' % ucsc_dbkey, 'chromFa', ucsc_dbkey]
255
256 ftp = FTP(UCSC_FTP_SERVER)
257 ftp.login('anonymous', email)
258
259 ucsc_path = UCSC_DOWNLOAD_PATH % ucsc_dbkey
260 path_contents = _get_files_in_ftp_path(ftp, ucsc_path)
261 ftp.quit()
262
263 for ucsc_chrom_fa_filename in UCSC_CHROM_FA_FILENAMES:
264 for ext in COMPRESSED_EXTENSIONS:
265 if "%s%s" % (ucsc_chrom_fa_filename, ext) in path_contents:
266 ucsc_file_name = "%s%s%s" % (ucsc_path, ucsc_chrom_fa_filename, ext)
267 return "ftp://%s%s" % (UCSC_FTP_SERVER, ucsc_file_name)
268
269 raise Exception('Unable to determine filename for UCSC Genome for %s: %s' % (ucsc_dbkey, path_contents))
270
271 def add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params):
272 for data_table_name, data_table_entry in _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params ):
273 if data_table_entry:
274 _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )
275
276
277 def download_from_ucsc( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
278 url = _get_ucsc_download_address(params, dbkey)
279 fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
280 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
281
282
283 def download_from_ncbi( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
284 NCBI_DOWNLOAD_URL = 'http://togows.dbcls.jp/entry/ncbi-nucleotide/%s.fasta' #FIXME: taken from dave's genome manager...why some japan site?
285 requested_identifier = params['param_dict']['reference_source']['requested_identifier']
286 url = NCBI_DOWNLOAD_URL % requested_identifier
287 fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
288 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
289
290
291 def download_from_url( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
292 urls = filter( bool, map( lambda x: x.strip(), params['param_dict']['reference_source']['user_url'].split( '\n' ) ) )
293 fasta_readers = [ get_stream_reader(urlopen( url ), tmp_dir) for url in urls ]
294 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id,sequence_name, params)
295
296
297 def download_from_history( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
298 #TODO: allow multiple FASTA input files
299 input_filename = params['param_dict']['reference_source']['input_fasta']
300 if isinstance( input_filename, list ):
301 fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
302 else:
303 fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
304 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
305
306
307 def copy_from_directory( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
308 input_filename = params['param_dict']['reference_source']['fasta_filename']
309 create_symlink = params['param_dict']['reference_source']['create_symlink'] == 'create_symlink'
310 if create_symlink:
311 data_table_entries = _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name )
312 else:
313 if isinstance( input_filename, list ):
314 fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
315 else:
316 fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
317 data_table_entries = _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params )
318 for data_table_name, data_table_entry in data_table_entries:
319 if data_table_entry:
320 _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )
321
322
323 def _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name ):
324 data_manager_dict['data_tables'] = data_manager_dict.get( 'data_tables', {} )
325 data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get( DATA_TABLE_NAME, [] )
326 data_manager_dict['data_tables'][data_table_name].append( data_table_entry )
327 return data_manager_dict
328
329
330 def _stream_fasta_to_file( fasta_stream, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params, close_stream=True ):
331 fasta_base_filename = "%s.gff" % sequence_id
332 fasta_filename = os.path.join( target_directory, fasta_base_filename )
333 with open( fasta_filename, 'wb+' ) as fasta_writer:
334
335 if isinstance( fasta_stream, list ) and len( fasta_stream ) == 1:
336 fasta_stream = fasta_stream[0]
337
338 if isinstance( fasta_stream, list ):
339 last_char = None
340 for fh in fasta_stream:
341 if last_char not in [ None, '\n', '\r', b'\n', b'\r' ]:
342 fasta_writer.write( b'\n' )
343 while True:
344 data = fh.read( CHUNK_SIZE )
345 if data:
346 fasta_writer.write( data )
347 last_char = data[-1]
348 else:
349 break
350 if close_stream:
351 fh.close()
352 else:
353 while True:
354 data = fasta_stream.read( CHUNK_SIZE )
355 if data:
356 fasta_writer.write( data )
357 else:
358 break
359 if close_stream:
360 fasta_stream.close()
361
362 #sort_fasta( fasta_filename, params['param_dict']['sorting']['sort_selector'], params )
363
364
365 return [ ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]
366
367
368 def compute_fasta_length( fasta_file, out_file, keep_first_word=False ):
369
370 infile = fasta_file
371 out = open( out_file, 'w')
372
373 fasta_title = ''
374 seq_len = 0
375
376 first_entry = True
377
378 for line in open( infile ):
379 line = line.strip()
380 if not line or line.startswith( '#' ):
381 continue
382 if line[0] == '>':
383 if first_entry == False:
384 if keep_first_word:
385 fasta_title = fasta_title.split()[0]
386 out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
387 else:
388 first_entry = False
389 fasta_title = line
390 seq_len = 0
391 else:
392 seq_len += len(line)
393
394 # last fasta-entry
395 if keep_first_word:
396 fasta_title = fasta_title.split()[0]
397 out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
398 out.close()
399
400
401 def _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name ):
402 fasta_base_filename = "%s.fa" % sequence_id
403 fasta_filename = os.path.join( target_directory, fasta_base_filename )
404 os.symlink( input_filename, fasta_filename )
405 return [ ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]
406
407
408 REFERENCE_SOURCE_TO_DOWNLOAD = dict( ucsc=download_from_ucsc, ncbi=download_from_ncbi, url=download_from_url, history=download_from_history, directory=copy_from_directory )
409
410 SORTING_METHODS = dict( as_is=_sort_fasta_as_is, lexicographical=_sort_fasta_lexicographical, gatk=_sort_fasta_gatk, custom=_sort_fasta_custom )
411
412
413 def main():
414 #Parse Command Line
415 parser = optparse.OptionParser()
416 parser.add_option( '-d', '--dbkey_description', dest='dbkey_description', action='store', type="string", default=None, help='dbkey_description' )
417 parser.add_option( '-t', '--type', dest='file_type', action='store', type='string', default=None, help='file_type')
418 (options, args) = parser.parse_args()
419
420 filename = args[0]
421 global DATA_TABLE_NAME
422 if options.file_type == 'representative':
423 DATA_TABLE_NAME= 'representative_gff'
424 params = loads( open( filename ).read() )
425 target_directory = params[ 'output_data' ][0]['extra_files_path']
426 os.mkdir( target_directory )
427 data_manager_dict = {}
428
429 dbkey, dbkey_name, sequence_id, sequence_name = get_dbkey_dbname_id_name( params, dbkey_description=options.dbkey_description )
430
431 if dbkey in [ None, '', '?' ]:
432 raise Exception( '"%s" is not a valid dbkey. You must specify a valid dbkey.' % ( dbkey ) )
433
434 # Create a tmp_dir, in case a zip file needs to be uncompressed
435 tmp_dir = tempfile.mkdtemp()
436 #Fetch the FASTA
437 try:
438 REFERENCE_SOURCE_TO_DOWNLOAD[ params['param_dict']['reference_source']['reference_source_selector'] ]( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir )
439 finally:
440 cleanup_before_exit(tmp_dir)
441 #save info to json file
442 open( filename, 'wb' ).write( dumps( data_manager_dict ).encode() )
443
444 if __name__ == "__main__":
445 main()