Mercurial > repos > ieguinoa > data_manager_fetch_gff

comparison data_manager/data_manager_fetch_gff.py @ 0:ac4fde07eaed draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author ieguinoa
date Mon, 09 Jul 2018 11:58:50 -0400
parents
children c57bd7f3fb46
comparison
equal deleted inserted replaced
-1:000000000000 0:ac4fde07eaed
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
55 sequence_name = params['param_dict']['sequence_name']
56 if not sequence_name:
57 sequence_name = dbkey_description
58 if not sequence_name:
59 sequence_name = dbkey
60 return dbkey, dbkey_name, sequence_id, sequence_name
61
62
63 def _get_files_in_ftp_path( ftp, path ):
64 path_contents = []
65 ftp.retrlines( 'MLSD %s' % ( path ), path_contents.append )
66 return [ line.split( ';' )[ -1 ].lstrip() for line in path_contents ]
67
68
69 def _get_stream_readers_for_tar( fh, tmp_dir ):
70 fasta_tar = tarfile.open( fileobj=fh, mode='r:*' )
71 return [x for x in [fasta_tar.extractfile(member) for member in fasta_tar.getmembers()] if x]
72
73
74 def _get_stream_readers_for_zip( fh, tmp_dir ):
75 """
76 Unpacks all archived files in a zip file.
77 Individual files will be concatenated (in _stream_fasta_to_file)
78 """
79 fasta_zip = zipfile.ZipFile( fh, 'r' )
80 rval = []
81 for member in fasta_zip.namelist():
82 fasta_zip.extract( member, tmp_dir )
83 rval.append( open( os.path.join( tmp_dir, member ), 'rb' ) )
84 return rval
85
86
87 def _get_stream_readers_for_gzip( fh, tmp_dir ):
88 return [ gzip.GzipFile( fileobj=fh, mode='rb') ]
89
90
91 def _get_stream_readers_for_bz2( fh, tmp_dir ):
92 return [ bz2.BZ2File( fh.name, 'rb') ]
93
94
95 def sort_fasta( fasta_filename, sort_method, params ):
96 if sort_method is None:
97 return
98 assert sort_method in SORTING_METHODS, ValueError( "%s is not a valid sorting option." % sort_method )
99 return SORTING_METHODS[ sort_method ]( fasta_filename, params )
100
101
102 def _move_and_index_fasta_for_sorting( fasta_filename ):
103 unsorted_filename = tempfile.NamedTemporaryFile().name
104 shutil.move( fasta_filename, unsorted_filename )
105 fasta_offsets = {}
106 unsorted_fh = open( unsorted_filename )
107 while True:
108 offset = unsorted_fh.tell()
109 line = unsorted_fh.readline()
110 if not line:
111 break
112 if line.startswith( ">" ):
113 line = line.split( None, 1 )[0][1:]
114 fasta_offsets[ line ] = offset
115 unsorted_fh.close()
116 current_order = map( lambda x: x[1], sorted( map( lambda x: ( x[1], x[0] ), fasta_offsets.items() ) ) )
117 return ( unsorted_filename, fasta_offsets, current_order )
118
119
120 def _write_sorted_fasta( sorted_names, fasta_offsets, sorted_fasta_filename, unsorted_fasta_filename ):
121 unsorted_fh = open( unsorted_fasta_filename )
122 sorted_fh = open( sorted_fasta_filename, 'wb+' )
123
124 for name in sorted_names:
125 offset = fasta_offsets[ name ]
126 unsorted_fh.seek( offset )
127 sorted_fh.write( unsorted_fh.readline() )
128 while True:
129 line = unsorted_fh.readline()
130 if not line or line.startswith( ">" ):
131 break
132 sorted_fh.write( line )
133 unsorted_fh.close()
134 sorted_fh.close()
135
136
137 def _sort_fasta_as_is( fasta_filename, params ):
138 return
139
140 def _sort_fasta_lexicographical( fasta_filename, params ):
141 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
142 sorted_names = sorted( fasta_offsets.keys() )
143 if sorted_names == current_order:
144 shutil.move( unsorted_filename, fasta_filename )
145 else:
146 _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
147
148
149 def _sort_fasta_gatk( fasta_filename, params ):
150 #This method was added by reviewer request.
151 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
152 sorted_names = map( str, range( 1, 23 ) ) + [ 'X', 'Y' ]
153 #detect if we have chrN, or just N
154 has_chr = False
155 for chrom in sorted_names:
156 if "chr%s" % chrom in current_order:
157 has_chr = True
158 break
159
160 if has_chr:
161 sorted_names = map( lambda x: "chr%s" % x, sorted_names)
162 sorted_names.insert( 0, "chrM" )
163 else:
164 sorted_names.insert( 0, "MT" )
165 sorted_names.extend( map( lambda x: "%s_random" % x, sorted_names ) )
166
167 existing_sorted_names = []
168 for name in sorted_names:
169 if name in current_order:
170 existing_sorted_names.append( name )
171 for name in current_order:
172 #TODO: confirm that non-canonical names do not need to be sorted specially
173 if name not in existing_sorted_names:
174 existing_sorted_names.append( name )
175
176 if existing_sorted_names == current_order:
177 shutil.move( unsorted_filename, fasta_filename )
178 else:
179 _write_sorted_fasta( existing_sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
180
181
182 def _sort_fasta_custom( fasta_filename, params ):
183 ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
184 sorted_names = []
185 for id_repeat in params['param_dict']['sorting']['sequence_identifiers']:
186 sorted_names.append( id_repeat[ 'identifier' ] )
187 handle_not_listed = params['param_dict']['sorting']['handle_not_listed_selector']
188 if handle_not_listed.startswith( 'keep' ):
189 add_list = []
190 for name in current_order:
191 if name not in sorted_names:
192 add_list.append( name )
193 if add_list:
194 if handle_not_listed == 'keep_append':
195 sorted_names.extend( add_list )
196 else:
197 add_list.extend( sorted_names )
198 sorted_names = add_list
199 if sorted_names == current_order:
200 shutil.move( unsorted_filename, fasta_filename )
201 else:
202 _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )
203
204
205 def _download_file(start, fh):
206 tmp = tempfile.NamedTemporaryFile()
207 tmp.write(start)
208 tmp.write(fh.read())
209 tmp.flush()
210 tmp.seek(0)
211 return tmp
212
213
214 def get_stream_reader(fh, tmp_dir):
215 """
216 Check if file is compressed and return correct stream reader.
217 If file has to be downloaded, do it now.
218 """
219 magic_dict = {
220 b"\x1f\x8b\x08": _get_stream_readers_for_gzip,
221 b"\x42\x5a\x68": _get_stream_readers_for_bz2,
222 b"\x50\x4b\x03\x04": _get_stream_readers_for_zip,
223 }
224 start_of_file = fh.read(CHUNK_SIZE)
225 try:
226 fh.seek(0)
227 except UnsupportedOperation: # This is if fh has been created by urlopen
228 fh = _download_file(start_of_file, fh)
229 for k,v in magic_dict.items():
230 if start_of_file.startswith(k):
231 return v(fh, tmp_dir)
232 try: # Check if file is tar file
233 if tarfile.open(fileobj=StringIO(start_of_file)):
234 return _get_stream_readers_for_tar(fh, tmp_dir)
235 except tarfile.ReadError:
236 pass
237 return fh
238
239
240 def _get_ucsc_download_address(params, dbkey):
241 """
242 Check if we can find the correct file for the supplied dbkey on UCSC's FTP server
243 """
244 UCSC_FTP_SERVER = 'hgdownload.cse.ucsc.edu'
245 UCSC_DOWNLOAD_PATH = '/goldenPath/%s/bigZips/'
246 COMPRESSED_EXTENSIONS = ['.tar.gz', '.tgz', '.tar.bz2', '.zip', '.fa.gz', '.fa.bz2']
247
248 email = params['param_dict']['__user_email__']
249 if not email:
250 email = 'anonymous@example.com'
251
252 ucsc_dbkey = params['param_dict']['reference_source']['requested_dbkey'] or dbkey
253 UCSC_CHROM_FA_FILENAMES = ['%s.chromFa' % ucsc_dbkey, 'chromFa', ucsc_dbkey]
254
255 ftp = FTP(UCSC_FTP_SERVER)
256 ftp.login('anonymous', email)
257
258 ucsc_path = UCSC_DOWNLOAD_PATH % ucsc_dbkey
259 path_contents = _get_files_in_ftp_path(ftp, ucsc_path)
260 ftp.quit()
261
262 for ucsc_chrom_fa_filename in UCSC_CHROM_FA_FILENAMES:
263 for ext in COMPRESSED_EXTENSIONS:
264 if "%s%s" % (ucsc_chrom_fa_filename, ext) in path_contents:
265 ucsc_file_name = "%s%s%s" % (ucsc_path, ucsc_chrom_fa_filename, ext)
266 return "ftp://%s%s" % (UCSC_FTP_SERVER, ucsc_file_name)
267
268 raise Exception('Unable to determine filename for UCSC Genome for %s: %s' % (ucsc_dbkey, path_contents))
269
270 def add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params):
271 for data_table_name, data_table_entry in _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params ):
272 if data_table_entry:
273 _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )
274
275
276 def download_from_ucsc( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
277 url = _get_ucsc_download_address(params, dbkey)
278 fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
279 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
280
281
282 def download_from_ncbi( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
283 NCBI_DOWNLOAD_URL = 'http://togows.dbcls.jp/entry/ncbi-nucleotide/%s.fasta' #FIXME: taken from dave's genome manager...why some japan site?
284 requested_identifier = params['param_dict']['reference_source']['requested_identifier']
285 url = NCBI_DOWNLOAD_URL % requested_identifier
286 fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
287 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
288
289
290 def download_from_url( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
291 urls = filter( bool, map( lambda x: x.strip(), params['param_dict']['reference_source']['user_url'].split( '\n' ) ) )
292 fasta_readers = [ get_stream_reader(urlopen( url ), tmp_dir) for url in urls ]
293 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id,sequence_name, params)
294
295
296 def download_from_history( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
297 #TODO: allow multiple FASTA input files
298 input_filename = params['param_dict']['reference_source']['input_fasta']
299 if isinstance( input_filename, list ):
300 fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
301 else:
302 fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
303 add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)
304
305
306 def copy_from_directory( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
307 input_filename = params['param_dict']['reference_source']['fasta_filename']
308 create_symlink = params['param_dict']['reference_source']['create_symlink'] == 'create_symlink'
309 if create_symlink:
310 data_table_entries = _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name )
311 else:
312 if isinstance( input_filename, list ):
313 fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
314 else:
315 fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
316 data_table_entries = _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params )
317 for data_table_name, data_table_entry in data_table_entries:
318 if data_table_entry:
319 _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )
320
321
322 def _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name ):
323 data_manager_dict['data_tables'] = data_manager_dict.get( 'data_tables', {} )
324 data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get( DATA_TABLE_NAME, [] )
325 data_manager_dict['data_tables'][data_table_name].append( data_table_entry )
326 return data_manager_dict
327
328
329 def _stream_fasta_to_file( fasta_stream, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params, close_stream=True ):
330 fasta_base_filename = "%s.gff" % sequence_id
331 fasta_filename = os.path.join( target_directory, fasta_base_filename )
332 with open( fasta_filename, 'wb+' ) as fasta_writer:
333
334 if isinstance( fasta_stream, list ) and len( fasta_stream ) == 1:
335 fasta_stream = fasta_stream[0]
336
337 if isinstance( fasta_stream, list ):
338 last_char = None
339 for fh in fasta_stream:
340 if last_char not in [ None, '\n', '\r', b'\n', b'\r' ]:
341 fasta_writer.write( b'\n' )
342 while True:
343 data = fh.read( CHUNK_SIZE )
344 if data:
345 fasta_writer.write( data )
346 last_char = data[-1]
347 else:
348 break
349 if close_stream:
350 fh.close()
351 else:
352 while True:
353 data = fasta_stream.read( CHUNK_SIZE )
354 if data:
355 fasta_writer.write( data )
356 else:
357 break
358 if close_stream:
359 fasta_stream.close()
360
361 #sort_fasta( fasta_filename, params['param_dict']['sorting']['sort_selector'], params )
362
363 dbkey_dict = None
364 if dbkey_name:
365 #do len calc here
366 #len_base_name = "%s.len" % ( dbkey )
367 #compute_fasta_length( fasta_filename, os.path.join( target_directory, len_base_name ), keep_first_word=True )
368 dbkey_dict = dict( value=dbkey, name=dbkey_name, len_path='' )
369
370 return [ ( '__dbkeys__', dbkey_dict ), ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]
371
372
373 def compute_fasta_length( fasta_file, out_file, keep_first_word=False ):
374
375 infile = fasta_file
376 out = open( out_file, 'w')
377
378 fasta_title = ''
379 seq_len = 0
380
381 first_entry = True
382
383 for line in open( infile ):
384 line = line.strip()
385 if not line or line.startswith( '#' ):
386 continue
387 if line[0] == '>':
388 if first_entry == False:
389 if keep_first_word:
390 fasta_title = fasta_title.split()[0]
391 out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
392 else:
393 first_entry = False
394 fasta_title = line
395 seq_len = 0
396 else:
397 seq_len += len(line)
398
399 # last fasta-entry
400 if keep_first_word:
401 fasta_title = fasta_title.split()[0]
402 out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
403 out.close()
404
405
406 def _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name ):
407 fasta_base_filename = "%s.fa" % sequence_id
408 fasta_filename = os.path.join( target_directory, fasta_base_filename )
409 os.symlink( input_filename, fasta_filename )
410
411 dbkey_dict = None
412 if dbkey_name:
413 #do len calc here
414 len_base_name = "%s.len" % ( dbkey )
415 compute_fasta_length( fasta_filename, os.path.join( target_directory, len_base_name ), keep_first_word=True )
416 dbkey_dict = dict( value=dbkey, name=dbkey_name, len_path=len_base_name )
417
418 return [ ( '__dbkeys__', dbkey_dict ), ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]
419
420
421 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 )
422
423 SORTING_METHODS = dict( as_is=_sort_fasta_as_is, lexicographical=_sort_fasta_lexicographical, gatk=_sort_fasta_gatk, custom=_sort_fasta_custom )
424
425
426 def main():
427 #Parse Command Line
428 parser = optparse.OptionParser()
429 parser.add_option( '-d', '--dbkey_description', dest='dbkey_description', action='store', type="string", default=None, help='dbkey_description' )
430 parser.add_option( '-t', '--type', dest='file_type', action='store', type='string', default=None, help='file_type')
431 (options, args) = parser.parse_args()
432
433 filename = args[0]
434 global DATA_TABLE_NAME
435 if options.file_type == 'representative':
436 DATA_TABLE_NAME= 'representative_gff'
437 params = loads( open( filename ).read() )
438 target_directory = params[ 'output_data' ][0]['extra_files_path']
439 os.mkdir( target_directory )
440 data_manager_dict = {}
441
442 dbkey, dbkey_name, sequence_id, sequence_name = get_dbkey_dbname_id_name( params, dbkey_description=options.dbkey_description )
443
444 if dbkey in [ None, '', '?' ]:
445 raise Exception( '"%s" is not a valid dbkey. You must specify a valid dbkey.' % ( dbkey ) )
446
447 # Create a tmp_dir, in case a zip file needs to be uncompressed
448 tmp_dir = tempfile.mkdtemp()
449 #Fetch the FASTA
450 try:
451 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 )
452 finally:
453 cleanup_before_exit(tmp_dir)
454 #save info to json file
455 open( filename, 'wb' ).write( dumps( data_manager_dict ).encode() )
456
457 if __name__ == "__main__":
458 main()