Mercurial > repos > xuebing > sharplabtool

comparison tools/extract/extract_genomic_dna.py @ 0:9071e359b9a3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author xuebing
date Fri, 09 Mar 2012 19:37:19 -0500
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:9071e359b9a3
1 #!/usr/bin/env python
2 """
3 usage: %prog $input $out_file1
4 -1, --cols=N,N,N,N: Columns for start, end, strand in input file
5 -d, --dbkey=N: Genome build of input file
6 -o, --output_format=N: the data type of the output file
7 -g, --GALAXY_DATA_INDEX_DIR=N: the directory containing alignseq.loc
8 -I, --interpret_features: if true, complete features are interpreted when input is GFF
9 -F, --fasta=<genomic_sequences>: genomic sequences to use for extraction
10 -G, --gff: input and output file, when it is interval, coordinates are treated as GFF format (1-based, half-open) rather than 'traditional' 0-based, closed format.
11 """
12 from galaxy import eggs
13 import pkg_resources
14 pkg_resources.require( "bx-python" )
15 import sys, string, os, re, tempfile, subprocess
16 from bx.cookbook import doc_optparse
17 from bx.intervals.io import Header, Comment
18 import bx.seq.nib
19 import bx.seq.twobit
20 from galaxy.tools.util.galaxyops import *
21 from galaxy.datatypes.util import gff_util
22
23 assert sys.version_info[:2] >= ( 2, 4 )
24
25 def stop_err( msg ):
26 sys.stderr.write( msg )
27 sys.exit()
28
29 def reverse_complement( s ):
30 complement_dna = {"A":"T", "T":"A", "C":"G", "G":"C", "a":"t", "t":"a", "c":"g", "g":"c", "N":"N", "n":"n" }
31 reversed_s = []
32 for i in s:
33 reversed_s.append( complement_dna[i] )
34 reversed_s.reverse()
35 return "".join( reversed_s )
36
37 def check_seq_file( dbkey, GALAXY_DATA_INDEX_DIR ):
38 seq_file = "%s/alignseq.loc" % GALAXY_DATA_INDEX_DIR
39 seq_path = ''
40 for line in open( seq_file ):
41 line = line.rstrip( '\r\n' )
42 if line and not line.startswith( "#" ) and line.startswith( 'seq' ):
43 fields = line.split( '\t' )
44 if len( fields ) < 3:
45 continue
46 if fields[1] == dbkey:
47 seq_path = fields[2].strip()
48 break
49 return seq_path
50
51 def __main__():
52 #
53 # Parse options, args.
54 #
55 options, args = doc_optparse.parse( __doc__ )
56 try:
57 chrom_col, start_col, end_col, strand_col = parse_cols_arg( options.cols )
58 dbkey = options.dbkey
59 output_format = options.output_format
60 gff_format = options.gff
61 interpret_features = options.interpret_features
62 GALAXY_DATA_INDEX_DIR = options.GALAXY_DATA_INDEX_DIR
63 fasta_file = options.fasta
64 input_filename, output_filename = args
65 except:
66 doc_optparse.exception()
67
68 includes_strand_col = strand_col >= 0
69 strand = None
70 nibs = {}
71 twobits = {}
72
73 #
74 # Set path to sequence data.
75 #
76 if fasta_file:
77 # Need to create 2bit file from fasta file.
78 try:
79 seq_path = tempfile.NamedTemporaryFile( dir="." ).name
80 cmd = "faToTwoBit %s %s" % ( fasta_file, seq_path )
81
82 tmp_name = tempfile.NamedTemporaryFile( dir="." ).name
83 tmp_stderr = open( tmp_name, 'wb' )
84 proc = subprocess.Popen( args=cmd, shell=True, stderr=tmp_stderr.fileno() )
85 returncode = proc.wait()
86 tmp_stderr.close()
87
88 # Get stderr, allowing for case where it's very large.
89 tmp_stderr = open( tmp_name, 'rb' )
90 stderr = ''
91 buffsize = 1048576
92 try:
93 while True:
94 stderr += tmp_stderr.read( buffsize )
95 if not stderr or len( stderr ) % buffsize != 0:
96 break
97 except OverflowError:
98 pass
99 tmp_stderr.close()
100
101 # Error checking.
102 if returncode != 0:
103 raise Exception, stderr
104 except Exception, e:
105 stop_err( 'Error running faToTwoBit. ' + str( e ) )
106 else:
107 seq_path = check_seq_file( dbkey, GALAXY_DATA_INDEX_DIR )
108 if not os.path.exists( seq_path ):
109 # If this occurs, we need to fix the metadata validator.
110 stop_err( "No sequences are available for '%s', request them by reporting this error." % dbkey )
111
112 #
113 # Fetch sequences.
114 #
115
116 # Get feature's line(s).
117 def get_lines( feature ):
118 if isinstance( feature, gff_util.GFFFeature ):
119 return feature.lines()
120 else:
121 return [ feature.rstrip( '\r\n' ) ]
122
123 skipped_lines = 0
124 first_invalid_line = 0
125 invalid_lines = []
126 fout = open( output_filename, "w" )
127 warnings = []
128 warning = ''
129 twobitfile = None
130 file_iterator = open( input_filename )
131 if gff_format and interpret_features:
132 file_iterator = gff_util.GFFReaderWrapper( file_iterator, fix_strand=False )
133 line_count = 1
134 for feature in file_iterator:
135 # Ignore comments, headers.
136 if isinstance( feature, ( Header, Comment ) ):
137 line_count += 1
138 continue
139
140 if gff_format and interpret_features:
141 # Processing features.
142 gff_util.convert_gff_coords_to_bed( feature )
143 chrom = feature.chrom
144 start = feature.start
145 end = feature.end
146 strand = feature.strand
147 else:
148 # Processing lines, either interval or GFF format.
149 line = feature.rstrip( '\r\n' )
150 if line and not line.startswith( "#" ):
151 fields = line.split( '\t' )
152 try:
153 chrom = fields[chrom_col]
154 start = int( fields[start_col] )
155 end = int( fields[end_col] )
156 if gff_format:
157 start, end = gff_util.convert_gff_coords_to_bed( [start, end] )
158 if includes_strand_col:
159 strand = fields[strand_col]
160 except:
161 warning = "Invalid chrom, start or end column values. "
162 warnings.append( warning )
163 if not invalid_lines:
164 invalid_lines = get_lines( feature )
165 first_invalid_line = line_count
166 skipped_lines += len( invalid_lines )
167 continue
168 if start > end:
169 warning = "Invalid interval, start '%d' > end '%d'. " % ( start, end )
170 warnings.append( warning )
171 if not invalid_lines:
172 invalid_lines = get_lines( feature )
173 first_invalid_line = line_count
174 skipped_lines += len( invalid_lines )
175 continue
176
177 if strand not in ['+', '-']:
178 strand = '+'
179 sequence = ''
180 else:
181 continue
182
183 # Open sequence file and get sequence for feature/interval.
184 if seq_path and os.path.exists( "%s/%s.nib" % ( seq_path, chrom ) ):
185 # TODO: improve support for GFF-nib interaction.
186 if chrom in nibs:
187 nib = nibs[chrom]
188 else:
189 nibs[chrom] = nib = bx.seq.nib.NibFile( file( "%s/%s.nib" % ( seq_path, chrom ) ) )
190 try:
191 sequence = nib.get( start, end-start )
192 except Exception, e:
193 warning = "Unable to fetch the sequence from '%d' to '%d' for build '%s'. " %( start, end-start, dbkey )
194 warnings.append( warning )
195 if not invalid_lines:
196 invalid_lines = get_lines( feature )
197 first_invalid_line = line_count
198 skipped_lines += len( invalid_lines )
199 continue
200 elif seq_path and os.path.isfile( seq_path ):
201 if not(twobitfile):
202 twobitfile = bx.seq.twobit.TwoBitFile( file( seq_path ) )
203 try:
204 if options.gff and interpret_features:
205 # Create sequence from intervals within a feature.
206 sequence = ''
207 for interval in feature.intervals:
208 sequence += twobitfile[interval.chrom][interval.start:interval.end]
209 else:
210 sequence = twobitfile[chrom][start:end]
211 except:
212 warning = "Unable to fetch the sequence from '%d' to '%d' for chrom '%s'. " %( start, end-start, chrom )
213 warnings.append( warning )
214 if not invalid_lines:
215 invalid_lines = get_lines( feature )
216 first_invalid_line = line_count
217 skipped_lines += len( invalid_lines )
218 continue
219 else:
220 warning = "Chromosome by name '%s' was not found for build '%s'. " % ( chrom, dbkey )
221 warnings.append( warning )
222 if not invalid_lines:
223 invalid_lines = get_lines( feature )
224 first_invalid_line = line_count
225 skipped_lines += len( invalid_lines )
226 continue
227 if sequence == '':
228 warning = "Chrom: '%s', start: '%s', end: '%s' is either invalid or not present in build '%s'. " \
229 % ( chrom, start, end, dbkey )
230 warnings.append( warning )
231 if not invalid_lines:
232 invalid_lines = get_lines( feature )
233 first_invalid_line = line_count
234 skipped_lines += len( invalid_lines )
235 continue
236 if includes_strand_col and strand == "-":
237 sequence = reverse_complement( sequence )
238
239 if output_format == "fasta" :
240 l = len( sequence )
241 c = 0
242 if gff_format:
243 start, end = gff_util.convert_bed_coords_to_gff( [ start, end ] )
244 fields = [dbkey, str( chrom ), str( start ), str( end ), strand]
245 meta_data = "_".join( fields )
246 fout.write( ">%s\n" % meta_data )
247 while c < l:
248 b = min( c + 50, l )
249 fout.write( "%s\n" % str( sequence[c:b] ) )
250 c = b
251 else: # output_format == "interval"
252 if gff_format and interpret_features:
253 # TODO: need better GFF Reader to capture all information needed
254 # to produce this line.
255 meta_data = "\t".join(
256 [feature.chrom, "galaxy_extract_genomic_dna", "interval", \
257 str( feature.start ), str( feature.end ), feature.score, feature.strand,
258 ".", gff_util.gff_attributes_to_str( feature.attributes, "GTF" ) ] )
259 else:
260 meta_data = "\t".join( fields )
261 if gff_format:
262 format_str = "%s seq \"%s\";\n"
263 else:
264 format_str = "%s\t%s\n"
265 fout.write( format_str % ( meta_data, str( sequence ) ) )
266
267 # Update line count.
268 if isinstance( feature, gff_util.GFFFeature ):
269 line_count += len( feature.intervals )
270 else:
271 line_count += 1
272
273 fout.close()
274
275 if warnings:
276 warn_msg = "%d warnings, 1st is: " % len( warnings )
277 warn_msg += warnings[0]
278 print warn_msg
279 if skipped_lines:
280 # Error message includes up to the first 10 skipped lines.
281 print 'Skipped %d invalid lines, 1st is #%d, "%s"' % ( skipped_lines, first_invalid_line, '\n'.join( invalid_lines[:10] ) )
282
283 if __name__ == "__main__": __main__()