opensource/galaxy
3
0
Fork 0
mirror of https://github.com/galaxyproject/galaxy.git synced 2026-09-21 13:50:20 +08:00
Code Issues Packages Projects Releases Wiki Activity

A tool for filtering SAM datasets on bitwise flag.

Browse Source
This commit is contained in:
Anton Nekrutenko
2009-09-16 15:19:35 -04:00
parent e6e5bd24b5
commit b833f1819d
2 changed files with 273 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/samtools/sam_bitwise_flag_filter.py
Executable
+177
View File
@@ -0,0 +1,177 @@
#!/usr/bin/env python
import sys
import optparse
def stop_err( msg ):
sys.stderr.write( msg )
sys.exit()
def main():
usage = """%prog [options]
options (listed below) default to 'None' if omitted
"""
parser = optparse.OptionParser(usage=usage)
parser.add_option(
'--0x0001','--is_paired',
choices = ( '0','1' ),
dest='is_paired',
metavar="<0|1>",
help='The read is paired in sequencing')
parser.add_option(
'--0x0002','--is_proper_pair',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_proper_pair',
help='The read is mapped in a proper pair')
parser.add_option(
'--0x0004','--is_unmapped',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_unmapped',
help='The query sequence itself is unmapped')
parser.add_option(
'--0x0008','--mate_is_unmapped',
choices = ( '0','1' ),
metavar="<0|1>",
dest='mate_is_unmapped',
help='The mate is unmapped')
parser.add_option(
'--0x0010','--query_strand',
dest='query_strand',
metavar="<0|1>",
choices = ( '0','1' ),
help='Strand of the query: 0 = forward, 1 = reverse.')
parser.add_option(
'--0x0020','--mate_strand',
dest='mate_strand',
metavar="<0|1>",
choices = ('0','1'),
help='Strand of the mate: 0 = forward, 1 = reverse.')
parser.add_option(
'--0x0040','--is_first',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_first',
help='The read is the first read in a pair')
parser.add_option(
'--0x0080','--is_second',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_second',
help='The read is the second read in a pair')
parser.add_option(
'--0x0100','--is_not_primary',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_not_primary',
help='The alignment for the given read is not primary')
parser.add_option(
'--0x0200','--is_bad_quality',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_bad_quality',
help='The read fails platform/vendor quality checks')
parser.add_option(
'--0x0400','--is_duplicate',
choices = ( '0','1' ),
metavar="<0|1>",
dest='is_duplicate',
help='The read is either a PCR or an optical duplicate')
parser.add_option(
'-f','--input_sam_file',
metavar="INPUT_SAM_FILE",
dest='input_sam',
default = False,
help='Name of the SAM file to be filtered. STDIN is default')
parser.add_option(
'-c','--flag_column',
dest='flag_col',
default = '2',
help='Column containing SAM bitwise flag. 1-based')
parser.add_option(
'-d','--debug',
dest='debug',
action='store_true',
default = False,
help='Print debugging info')
options, args = parser.parse_args()
if options.input_sam:
infile = open ( options.input_sam, 'r')
else:
infile = sys.stdin
option_values = { '0': False, '1': True, None: None }
states = [];
states.append( option_values[ options.is_paired ] )
states.append( option_values[ options.is_proper_pair ] )
states.append( option_values[ options.is_unmapped ] )
states.append( option_values[ options.mate_is_unmapped ] )
states.append( option_values[ options.query_strand ] )
states.append( option_values[ options.mate_strand ] )
states.append( option_values[ options.is_first ] )
states.append( option_values[ options.is_second ] )
states.append( option_values[ options.is_not_primary ] )
states.append( option_values[ options.is_bad_quality ] )
states.append( option_values[ options.is_duplicate ] )
for line in infile:
line = line.rstrip( '\r\n' )
if line and not line.startswith( '#' ):
fields = line.split( '\t' )
sam_states = []
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0001 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0002 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0004 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0008 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0010 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0020 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0040 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0080 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0100 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0200 ) )
sam_states.append( bool( int( fields[ int( options.flag_col ) - 1 ] ) & 0x0400 ) )
joined_states = zip(states,sam_states)
searchable_fields = []
for i in range( len( joined_states ) ):
if joined_states[i][0] != None:
searchable_fields.append( joined_states[ i ] )
valid_line = True
for i in range( len( searchable_fields ) ):
if searchable_fields[i][0] != searchable_fields[i][1]:
valid_line = False
if valid_line:
print line
if options.debug:
for i in range( len( joined_states ) ):
print i, joined_states[i][0], joined_states[i][1]
# if skipped_lines > 0:
# print 'Skipped %d invalid lines' % skipped_lines
if __name__ == "__main__": main()
tools/samtools/sam_bitwise_flag_filter.xml
+96
View File
@@ -0,0 +1,96 @@
<tool id="sam_bw_filter" name="Filter SAM" version="1.0.0">
<description>on bitwise flag values</description>
<command interpreter="python">
sam_bitwise_flag_filter.py
--input_sam_file=$input1
--flag_column=2
#for $bit in $bits
'${bit.flags}=${bit.states}'
#end for
> $out_file1
</command>
<inputs>
<param format="sam" name="input1" type="data" label="Select dataset to filter"/>
<repeat name="bits" title="Flag">
<param name="flags" type="select" label="Type">
<option value="--0x0001">Read is paired</option>
<option value="--0x0002">Read is mapped in a proper pair</option>
<option value="--0x0004">The read is unmapped</option>
<option value="--0x0008">The mate is unmapped</option>
<option value="--0x0010">Read strand</option>
<option value="--0x0020">Mate strand</option>
<option value="--0x0040">Read is the first in a pair</option>
<option value="--0x0080">Read is the second in a pair</option>
<option value="--0x0100">The alignment or this read is not primary</option>
<option value="--0x0200">The read fails platform/vendor quality checks</option>
<option value="--0x0400">The read is a PCR or optical duplicate</option>
</param>
<param name="states" type="select" display="radio" label="Set the states for this flag">
<option value="0">No</option>
<option value="1">Yes</option>
</param>
</repeat>
</inputs>
<outputs>
<data format="sam" name="out_file1" />
</outputs>
<tests>
<test>
<param name="input1" value="sam_bw_filter.sam" ftype="sam"/>
<param name="flags" value="Read is mapped in a proper pair"/>
<param name="states" value="1"/>
<output name="out_file1" file="sam_bw_filter_0002-yes.sam" ftype="sam"/>
</test>
</tests>
<help>
**What it does**
Allows parsing SAM datasets using bitwise flag (the second column). The bits in the flag are defined as follows::
Bit Info
------ --------------------------------------------------------------------------
0x0001 the read is paired in sequencing, no matter whether it is mapped in a pair
0x0002 the read is mapped in a proper pair (depends on the protocol, normally
inferred during alignment) 1
0x0004 the query sequence itself is unmapped
0x0008 the mate is unmapped 1
0x0010 strand of the query (0 for forward; 1 for reverse strand)
0x0020 strand of the mate 1
0x0040 the read is the first read in a pair (see below)
0x0080 the read is the second read in a pair (see below)
0x0100 the alignment is not primary (a read having split hits may
have multiple primary alignment records)
0x0200 the read fails platform/vendor quality checks
0x0400 the read is either a PCR duplicate or an optical duplicate
Note the following:
- Flag 0x02, 0x08, 0x20, 0x40 and 0x80 are only meaningful when flag 0x01 is present.
- If in a read pair the information on which read is the first in the pair is lost in the upstream analysis, flag 0x01 should be present and 0x40 and 0x80 are both zero.
-----
**Example**
Suppose the following dataset was generated with BWA mapper::
r001 163 ref 7 30 8M2I4M1D3M = 37 39 TTAGATAAAGGATACTA *
r002 0 ref 9 30 3S6M1P1I4M * 0 0 AAAAGATAAGGATA *
r003 0 ref 9 30 5H6M * 0 0 AGCTAA * NM:i:1
r004 0 ref 16 30 6M14N5M * 0 0 ATAGCTTCAGC *
r003 16 ref 29 30 6H5M * 0 0 TAGGC * NM:i:0
r001 83 ref 37 30 9M = 7 -39 CAGCGCCAT *
To select properly mapped pairs click the **Add new Flag** button and set *Read mapped in a proper pair* to **Yes**. The following two reads will be returned::
r001 163 ref 7 30 8M2I4M1D3M = 37 39 TTAGATAAAGGATACTA *
r001 83 ref 37 30 9M = 7 -39 CAGCGCCAT *
For more information please consult the `SAM format description`__.
.. __: http://www.ncbi.nlm.nih.gov/pubmed/19505943
</help>
</tool>