|
0
|
1 #!/usr/bin/env python
|
|
|
2 """
|
|
|
3 Cluster regions of intervals.
|
|
|
4
|
|
|
5 usage: %prog in_file out_file
|
|
|
6 -1, --cols1=N,N,N,N: Columns for start, end, strand in file
|
|
|
7 -d, --distance=N: Maximum distance between clustered intervals
|
|
|
8 -v, --overlap=N: Minimum overlap require (negative distance)
|
|
|
9 -m, --minregions=N: Minimum regions per cluster
|
|
|
10 -o, --output=N: 1)merged 2)filtered 3)clustered 4) minimum 5) maximum
|
|
|
11 """
|
|
|
12 import sys, traceback, fileinput
|
|
|
13 from warnings import warn
|
|
|
14 from bx.intervals import *
|
|
|
15 from bx.intervals.io import *
|
|
|
16 from bx.intervals.operations.find_clusters import *
|
|
|
17 from bx.cookbook import doc_optparse
|
|
|
18 from galaxy.tools.util.galaxyops import *
|
|
|
19
|
|
|
20 assert sys.version_info[:2] >= ( 2, 4 )
|
|
|
21
|
|
|
22 def main():
|
|
|
23 distance = 0
|
|
|
24 minregions = 2
|
|
|
25 output = 1
|
|
|
26 upstream_pad = 0
|
|
|
27 downstream_pad = 0
|
|
|
28
|
|
|
29 options, args = doc_optparse.parse( __doc__ )
|
|
|
30 try:
|
|
|
31 chr_col_1, start_col_1, end_col_1, strand_col_1 = parse_cols_arg( options.cols1 )
|
|
|
32 if options.distance: distance = int( options.distance )
|
|
|
33 if options.overlap: distance = -1 * int( options.overlap )
|
|
|
34 if options.output: output = int( options.output )
|
|
|
35 if options.minregions: minregions = int( options.minregions )
|
|
|
36 in_fname, out_fname = args
|
|
|
37 except:
|
|
|
38 doc_optparse.exception()
|
|
|
39
|
|
|
40 g1 = NiceReaderWrapper( fileinput.FileInput( in_fname ),
|
|
|
41 chrom_col=chr_col_1,
|
|
|
42 start_col=start_col_1,
|
|
|
43 end_col=end_col_1,
|
|
|
44 strand_col=strand_col_1,
|
|
|
45 fix_strand=True )
|
|
|
46
|
|
|
47 # Get the cluster tree
|
|
|
48 try:
|
|
|
49 clusters, extra = find_clusters( g1, mincols=distance, minregions=minregions)
|
|
|
50 except ParseError, exc:
|
|
|
51 fail( "Invalid file format: %s" % str( exc ) )
|
|
|
52
|
|
|
53 f1 = open( in_fname, "r" )
|
|
|
54 out_file = open( out_fname, "w" )
|
|
|
55
|
|
|
56 # If "merge"
|
|
|
57 if output == 1:
|
|
|
58 fields = ["." for x in range(max(g1.chrom_col, g1.start_col, g1.end_col)+1)]
|
|
|
59 for chrom, tree in clusters.items():
|
|
|
60 for start, end, lines in tree.getregions():
|
|
|
61 fields[g1.chrom_col] = chrom
|
|
|
62 fields[g1.start_col] = str(start)
|
|
|
63 fields[g1.end_col] = str(end)
|
|
|
64 out_file.write( "%s\n" % "\t".join( fields ) )
|
|
|
65
|
|
|
66 # If "filtered" we preserve order of file and comments, etc.
|
|
|
67 if output == 2:
|
|
|
68 linenums = dict()
|
|
|
69 for chrom, tree in clusters.items():
|
|
|
70 for linenum in tree.getlines():
|
|
|
71 linenums[linenum] = 0
|
|
|
72 linenum = -1
|
|
|
73 f1.seek(0)
|
|
|
74 for line in f1.readlines():
|
|
|
75 linenum += 1
|
|
|
76 if linenum in linenums or linenum in extra:
|
|
|
77 out_file.write( "%s\n" % line.rstrip( "\n\r" ) )
|
|
|
78
|
|
|
79 # If "clustered" we output original intervals, but near each other (i.e. clustered)
|
|
|
80 if output == 3:
|
|
|
81 linenums = list()
|
|
|
82 f1.seek(0)
|
|
|
83 fileLines = f1.readlines()
|
|
|
84 for chrom, tree in clusters.items():
|
|
|
85 for linenum in tree.getlines():
|
|
|
86 out_file.write( "%s\n" % fileLines[linenum].rstrip( "\n\r" ) )
|
|
|
87
|
|
|
88 # If "minimum" we output the smallest interval in each cluster
|
|
|
89 if output == 4 or output == 5:
|
|
|
90 linenums = list()
|
|
|
91 f1.seek(0)
|
|
|
92 fileLines = f1.readlines()
|
|
|
93 for chrom, tree in clusters.items():
|
|
|
94 regions = tree.getregions()
|
|
|
95 for start, end, lines in tree.getregions():
|
|
|
96 outsize = -1
|
|
|
97 outinterval = None
|
|
|
98 for line in lines:
|
|
|
99 # three nested for loops?
|
|
|
100 # should only execute this code once per line
|
|
|
101 fileline = fileLines[line].rstrip("\n\r")
|
|
|
102 try:
|
|
|
103 cluster_interval = GenomicInterval( g1, fileline.split("\t"),
|
|
|
104 g1.chrom_col,
|
|
|
105 g1.start_col,
|
|
|
106 g1.end_col,
|
|
|
107 g1.strand_col,
|
|
|
108 g1.default_strand,
|
|
|
109 g1.fix_strand )
|
|
|
110 except Exception, exc:
|
|
|
111 print >> sys.stderr, str( exc )
|
|
|
112 f1.close()
|
|
|
113 sys.exit()
|
|
|
114 interval_size = cluster_interval.end - cluster_interval.start
|
|
|
115 if outsize == -1 or \
|
|
|
116 ( outsize > interval_size and output == 4 ) or \
|
|
|
117 ( outsize < interval_size and output == 5 ) :
|
|
|
118 outinterval = cluster_interval
|
|
|
119 outsize = interval_size
|
|
|
120 out_file.write( "%s\n" % outinterval )
|
|
|
121
|
|
|
122 f1.close()
|
|
|
123 out_file.close()
|
|
|
124
|
|
|
125 if g1.skipped > 0:
|
|
|
126 print skipped( g1, filedesc="" )
|
|
|
127
|
|
|
128 if __name__ == "__main__":
|
|
|
129 main()
|
