|
0
|
1 #purpose: macs2 python wrapper
|
|
|
2 #author: Ziru Zhou
|
|
|
3 #date: November, 2012
|
|
|
4
|
|
|
5 import sys, subprocess, tempfile, shutil, glob, os, os.path, gzip
|
|
|
6 from galaxy import eggs
|
|
|
7 import pkg_resources
|
|
|
8 pkg_resources.require( "simplejson" )
|
|
|
9 import simplejson
|
|
|
10
|
|
|
11 CHUNK_SIZE = 1024
|
|
|
12
|
|
|
13 #==========================================================================================
|
|
|
14 #functions
|
|
|
15 #==========================================================================================
|
|
|
16 def gunzip_cat_glob_path( glob_path, target_filename, delete = False ):
|
|
|
17 out = open( target_filename, 'wb' )
|
|
|
18 for filename in glob.glob( glob_path ):
|
|
|
19 fh = gzip.open( filename, 'rb' )
|
|
|
20 while True:
|
|
|
21 data = fh.read( CHUNK_SIZE )
|
|
|
22 if data:
|
|
|
23 out.write( data )
|
|
|
24 else:
|
|
|
25 break
|
|
|
26 fh.close()
|
|
|
27 if delete:
|
|
|
28 os.unlink( filename )
|
|
|
29 out.close()
|
|
|
30
|
|
|
31 def xls_to_interval( xls_file, interval_file, header = None ):
|
|
|
32 out = open( interval_file, 'wb' )
|
|
|
33 if header:
|
|
|
34 out.write( '#%s\n' % header )
|
|
|
35 wrote_header = False
|
|
|
36 #From macs readme: Coordinates in XLS is 1-based which is different with BED format.
|
|
|
37 for line in open( xls_file ):
|
|
|
38 #keep all existing comment lines
|
|
|
39 if line.startswith( '#' ):
|
|
|
40 out.write( line )
|
|
|
41 #added for macs2 since there is an extra newline
|
|
|
42 elif line.startswith( '\n' ):
|
|
|
43 out.write( line )
|
|
|
44 elif not wrote_header:
|
|
|
45 out.write( '#%s' % line )
|
|
|
46 print line
|
|
|
47 wrote_header = True
|
|
|
48 else:
|
|
|
49 fields = line.split( '\t' )
|
|
|
50 if len( fields ) > 1:
|
|
|
51 fields[1] = str( int( fields[1] ) - 1 )
|
|
|
52 out.write( '\t'.join( fields ) )
|
|
|
53 out.close()
|
|
|
54
|
|
|
55 #==========================================================================================
|
|
|
56 #main
|
|
|
57 #==========================================================================================
|
|
|
58 def main():
|
|
|
59 #take in options file and output file names
|
|
|
60 options = simplejson.load( open( sys.argv[1] ) )
|
|
|
61 outputs = simplejson.load( open( sys.argv[2] ) )
|
|
|
62
|
|
|
63 #=================================================================================
|
|
|
64 #parse options and execute macs2
|
|
|
65 #=================================================================================
|
|
|
66 #default inputs that are in every major command
|
|
|
67 experiment_name = '_'.join( options['experiment_name'].split() ) #save experiment name here, it will be used by macs for some file names
|
|
|
68 cmdline = "macs2 %s -t %s" % ( options['command'], ",".join( options['input_chipseq'] ) )
|
|
|
69 if options['input_control']:
|
|
|
70 cmdline = "%s -c %s" % ( cmdline, ",".join( options['input_control'] ) )
|
|
|
71
|
|
|
72 #=================================================================================
|
|
|
73 if (options['command'] == "callpeak"):
|
|
|
74 output_bed = outputs['output_bed_file']
|
|
|
75 output_extra_html = outputs['output_extra_file']
|
|
|
76 output_extra_path = outputs['output_extra_file_path']
|
|
|
77 output_peaks = outputs['output_peaks_file']
|
|
|
78 output_narrowpeaks = outputs['output_narrowpeaks_file']
|
|
|
79 output_xls_to_interval_peaks_file = outputs['output_xls_to_interval_peaks_file']
|
|
|
80 output_xls_to_interval_negative_peaks_file = outputs['output_xls_to_interval_negative_peaks_file']
|
|
|
81
|
|
|
82 if 'pvalue' in options:
|
|
|
83 cmdline = "%s --format='%s' --name='%s' --gsize='%s' --bw='%s' --pvalue='%s' --mfold %s %s %s %s" % ( cmdline, options['format'], experiment_name, options['gsize'], options['bw'], options['pvalue'], options['mfoldlo'], options['mfoldhi'], options['nolambda'], options['bdg'] )
|
|
|
84 elif 'qvalue' in options:
|
|
|
85 cmdline = "%s --format='%s' --name='%s' --gsize='%s' --bw='%s' --qvalue='%s' --mfold %s %s %s %s" % ( cmdline, options['format'], experiment_name, options['gsize'], options['bw'], options['qvalue'], options['mfoldlo'], options['mfoldhi'], options['nolambda'], options['bdg'] )
|
|
|
86
|
|
|
87 if 'nomodel' in options:
|
|
|
88 cmdline = "%s --nomodel --shiftsize='%s'" % ( cmdline, options['nomodel'] )
|
|
|
89 #=================================================================================
|
|
|
90 if (options['command'] == "bdgcmp"):
|
|
|
91 output_bdgcmp = outputs['output_bdgcmp_file']
|
|
|
92
|
|
|
93 cmdline = "%s -m %s -p %s -o bdgcmp_out.bdg" % ( cmdline, options['m'], options['pseudocount'] )
|
|
|
94 #=================================================================================
|
|
|
95
|
|
|
96 tmp_dir = tempfile.mkdtemp() #macs makes very messy output, need to contain it into a temp dir, then provide to user
|
|
|
97 stderr_name = tempfile.NamedTemporaryFile().name # redirect stderr here, macs provides lots of info via stderr, make it into a report
|
|
|
98 proc = subprocess.Popen( args=cmdline, shell=True, cwd=tmp_dir, stderr=open( stderr_name, 'wb' ) )
|
|
|
99 proc.wait()
|
|
|
100 #We don't want to set tool run to error state if only warnings or info, e.g. mfold could be decreased to improve model, but let user view macs log
|
|
|
101 #Do not terminate if error code, allow dataset (e.g. log) creation and cleanup
|
|
|
102 if proc.returncode:
|
|
|
103 stderr_f = open( stderr_name )
|
|
|
104 while True:
|
|
|
105 chunk = stderr_f.read( CHUNK_SIZE )
|
|
|
106 if not chunk:
|
|
|
107 stderr_f.close()
|
|
|
108 break
|
|
|
109 sys.stderr.write( chunk )
|
|
|
110
|
|
|
111 #=================================================================================
|
|
|
112 #copy files created by macs2 to appripriate directory with the provided names
|
|
|
113 #=================================================================================
|
|
|
114
|
|
|
115 #=================================================================================
|
|
|
116 #move files generated by callpeak command
|
|
|
117 if (options['command'] == "callpeak"):
|
|
|
118 #run R to create pdf from model script
|
|
|
119 if os.path.exists( os.path.join( tmp_dir, "%s_model.r" % experiment_name ) ):
|
|
|
120 cmdline = 'R --vanilla --slave < "%s_model.r" > "%s_model.r.log"' % ( experiment_name, experiment_name )
|
|
|
121 proc = subprocess.Popen( args=cmdline, shell=True, cwd=tmp_dir )
|
|
|
122 proc.wait()
|
|
|
123
|
|
|
124 #move bed out to proper output file
|
|
|
125 created_bed_name = os.path.join( tmp_dir, "%s_peaks.bed" % experiment_name )
|
|
|
126 if os.path.exists( created_bed_name ):
|
|
|
127 shutil.move( created_bed_name, output_bed )
|
|
|
128
|
|
|
129 #OICR peak_xls file
|
|
|
130 created_peak_xls_file = os.path.join( tmp_dir, "%s_peaks.xls" % experiment_name )
|
|
|
131 if os.path.exists( created_peak_xls_file ):
|
|
|
132 # shutil.copy( created_peak_xls_file, os.path.join ( "/mnt/galaxyData/tmp/", "%s_peaks.xls" % ( os.path.basename(output_extra_path) )))
|
|
|
133 shutil.copyfile( created_peak_xls_file, output_peaks )
|
|
|
134
|
|
|
135 #peaks.encodepeaks (narrowpeaks) file
|
|
|
136 created_narrowpeak_file = os.path.join (tmp_dir, "%s_peaks.encodePeak" % experiment_name )
|
|
|
137 if os.path.exists( created_narrowpeak_file ):
|
|
|
138 shutil.move (created_narrowpeak_file, output_narrowpeaks )
|
|
|
139
|
|
|
140 #parse xls files to interval files as needed
|
|
|
141 #if 'xls_to_interval' in options:
|
|
|
142 if (options['xls_to_interval'] == "True"):
|
|
|
143 create_peak_xls_file = os.path.join( tmp_dir, '%s_peaks.xls' % experiment_name )
|
|
|
144 if os.path.exists( create_peak_xls_file ):
|
|
|
145 xls_to_interval( create_peak_xls_file, output_xls_to_interval_peaks_file, header = 'peaks file' )
|
|
|
146 create_peak_xls_file = os.path.join( tmp_dir, '%s_negative_peaks.xls' % experiment_name )
|
|
|
147 if os.path.exists( create_peak_xls_file ):
|
|
|
148 print "negative file exists"
|
|
|
149 xls_to_interval( create_peak_xls_file, output_xls_to_interval_negative_peaks_file, header = 'negative peaks file' )
|
|
|
150
|
|
|
151 #move all remaining files to extra files path of html file output to allow user download
|
|
|
152 out_html = open( output_extra_html, 'wb' )
|
|
|
153 out_html.write( '<html><head><title>Additional output created by MACS (%s)</title></head><body><h3>Additional Files:</h3><p><ul>\n' % experiment_name )
|
|
|
154 os.mkdir( output_extra_path )
|
|
|
155 for filename in sorted( os.listdir( tmp_dir ) ):
|
|
|
156 shutil.move( os.path.join( tmp_dir, filename ), os.path.join( output_extra_path, filename ) )
|
|
|
157 out_html.write( '<li><a href="%s">%s</a></li>\n' % ( filename, filename ) )
|
|
|
158 #out_html.write( '<li><a href="%s">%s</a>peakxls %s SomethingDifferent tmp_dir %s path %s exp_name %s</li>\n' % ( created_peak_xls_file, filename, filename, tmp_dir, output_extra_path, experiment_name ) )
|
|
|
159 out_html.write( '</ul></p>\n' )
|
|
|
160 out_html.write( '<h3>Messages from MACS:</h3>\n<p><pre>%s</pre></p>\n' % open( stderr_name, 'rb' ).read() )
|
|
|
161 out_html.write( '</body></html>\n' )
|
|
|
162 out_html.close()
|
|
|
163
|
|
|
164 #=================================================================================
|
|
|
165 #move files generated by bdgcmp command
|
|
|
166 if (options['command'] == "bdgcmp"):
|
|
|
167 created_bdgcmp_file = os.path.join (tmp_dir, "bdgcmp_out.bdg" )
|
|
|
168 if os.path.exists( created_bdgcmp_file ):
|
|
|
169 shutil.move (created_bdgcmp_file, output_bdgcmp )
|
|
|
170
|
|
|
171 #=================================================================================
|
|
|
172 #cleanup
|
|
|
173 #=================================================================================
|
|
|
174 os.unlink( stderr_name )
|
|
|
175 os.rmdir( tmp_dir )
|
|
|
176
|
|
|
177 if __name__ == "__main__": main()
|
