opensource/galaxy
3
0
Fork 0
mirror of https://github.com/galaxyproject/galaxy.git synced 2026-09-24 16:30:27 +08:00
Code Issues Packages Projects Releases Wiki Activity

Use GFFReaderWrapper in gff_filter_by_feature_count tool in order to leverage support for GTF, GFF, and GFF3 found in reader wrapper. This also simplifies the tool considerably.

Browse Source
This commit is contained in:
Jeremy Goecks
2011-03-08 19:01:59 -05:00
parent a4f910c730
commit 86b7194bff
2 changed files with 36 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/galaxy/datatypes/util/gff_util.py
+22 -16
View File
@@ -10,14 +10,18 @@ class GFFInterval( GenomicInterval ):
A GFF interval, including attributes. If file is strictly a GFF file,
only attribute is 'group.'
"""
def __init__( self, reader, fields, chrom_col, start_col, end_col, strand_col, \
score_col, default_strand, fix_strand=False, raw_line='' ):
def __init__( self, reader, fields, chrom_col, feature_col, start_col, end_col, \
strand_col, score_col, default_strand, fix_strand=False, raw_line='' ):
GenomicInterval.__init__( self, reader, fields, chrom_col, start_col, end_col, strand_col, \
default_strand, fix_strand=fix_strand )
# Handle score column.
# Handle feature, score column.
self.feature_col = feature_col
if self.feature_col >= self.nfields:
raise MissingFieldError( "No field for feature_col (%d)" % feature_col )
self.feature = self.fields[ self.feature_col ]
self.score_col = score_col
if self.score_col >= self.nfields:
raise MissingFieldError( "No field for score_col (%d)" % score_col )
raise MissingFieldError( "No field for score_col (%d)" % score_col )
self.score = self.fields[ self.score_col ]
# Attributes specific to GFF.
@@ -28,10 +32,11 @@ class GFFFeature( GFFInterval ):
"""
A GFF feature, which can include multiple intervals.
"""
def __init__( self, reader, chrom_col, start_col, end_col, strand_col, score_col, default_strand, \
fix_strand=False, intervals=[] ):
GFFInterval.__init__( self, reader, intervals[0].fields, chrom_col, start_col, end_col, \
strand_col, score_col, default_strand, fix_strand=fix_strand )
def __init__( self, reader, chrom_col, feature_col, start_col, end_col, \
strand_col, score_col, default_strand, fix_strand=False, intervals=[] ):
GFFInterval.__init__( self, reader, intervals[0].fields, chrom_col, feature_col, \
start_col, end_col, strand_col, score_col, default_strand, \
fix_strand=fix_strand )
self.intervals = intervals
# Use intervals to set feature attributes.
for interval in self.intervals:
@@ -99,20 +104,20 @@ class GFFReaderWrapper( NiceReaderWrapper ):
expect traditional interval format.
"""
def __init__( self, reader, chrom_col=0, start_col=3, end_col=4, strand_col=6, score_col=5, \
fix_strand=False, **kwargs ):
def __init__( self, reader, chrom_col=0, feature_col=2, start_col=3, \
end_col=4, strand_col=6, score_col=5, fix_strand=False, **kwargs ):
NiceReaderWrapper.__init__( self, reader, chrom_col=chrom_col, start_col=start_col, end_col=end_col, \
strand_col=strand_col, fix_strand=fix_strand, **kwargs )
# HACK: NiceReaderWrapper (bx-python) does not handle score_col yet, so store ourselves.
self.feature_col = feature_col
self.score_col = score_col
self.last_line = None
self.cur_offset = 0
self.seed_interval = None
def parse_row( self, line ):
interval = GFFInterval( self, line.split( "\t" ), self.chrom_col, self.start_col, \
self.end_col, self.strand_col, self.score_col, self.default_strand, \
fix_strand=self.fix_strand, raw_line=line )
interval = GFFInterval( self, line.split( "\t" ), self.chrom_col, self.feature_col, \
self.start_col, self.end_col, self.strand_col, self.score_col, \
self.default_strand, fix_strand=self.fix_strand, raw_line=line )
return interval
def next( self ):
@@ -196,8 +201,9 @@ class GFFReaderWrapper( NiceReaderWrapper ):
self.seed_interval = interval
# Return GFF feature with all intervals.
return GFFFeature( self, self.chrom_col, self.start_col, self.end_col, self.strand_col, \
self.score_col, self.default_strand, fix_strand=self.fix_strand, \
return GFFFeature( self, self.chrom_col, self.feature_col, self.start_col, \
self.end_col, self.strand_col, self.score_col, \
self.default_strand, fix_strand=self.fix_strand, \
intervals=feature_intervals )
tools/filters/gff/gff_filter_by_feature_count.py
+14 -66
View File
@@ -7,7 +7,7 @@ Usage:
"""
import sys
from galaxy import eggs
from galaxy.datatypes.util.gff_util import parse_gff_attributes
from galaxy.datatypes.util.gff_util import GFFReaderWrapper
assert sys.version_info[:2] >= ( 2, 4 )
@@ -58,77 +58,25 @@ def __main__():
break
# Do filtering.
kept_lines = 0
kept_features = 0
skipped_lines = 0
first_skipped_line = 0
out = open( output_name, 'w' )
i = 0
cur_transcript_id = None
cur_transcript_lines = []
cur_transcript_feature_counts = {} # Key is feature name, value is feature count.
for i, line in enumerate( file( input_name ) ):
line = line.rstrip( '\r\n' )
if line and not line.startswith( '#' ):
try:
# GFF format: chrom, source, feature, chromStart, chromEnd, score, strand, attributes
elems = line.split( '\t' )
feature = elems[2]
start = str( long( elems[3] ) - 1 )
coords = [ long( start ), long( elems[4] ) ]
strand = elems[6]
attributes = parse_gff_attributes( elems[8] )
t_id = attributes.get( "transcript_id", None )
if not t_id:
# No transcript id, so pass line to output.
out.write( line )
kept_lines += 1
continue
# There is a transcript ID, so process line at transcript level.
if t_id == cur_transcript_id:
# Line is element of transcript; increment feature count.
if not feature in cur_transcript_feature_counts:
cur_transcript_feature_counts[feature] = 0
cur_transcript_feature_counts[feature] += 1
cur_transcript_lines.append( line )
continue
#
# Line is part of new transcript; filter previous transcript.
#
# Filter/write previous transcript.
result = eval( '%s %s' % ( cur_transcript_feature_counts.get( feature_name, 0 ), condition ) )
if cur_transcript_id and result:
# Transcript passes filter; write transcript line to file."
out.write( "\n".join( cur_transcript_lines ) + "\n" )
kept_lines += len( cur_transcript_lines )
# Start new transcript.
cur_transcript_id = t_id
cur_transcript_feature_counts = {}
cur_transcript_feature_counts[feature] = 1
cur_transcript_lines = [ line ]
except Exception, e:
print e
skipped_lines += 1
if not first_skipped_line:
first_skipped_line = i + 1
else:
skipped_lines += 1
if not first_skipped_line:
first_skipped_line = i + 1
# Write last transcript.
if cur_transcript_id and eval( '%s %s' % ( cur_transcript_feature_counts[feature_name], condition ) ):
# Transcript passes filter; write transcript lints to file.
out.write( "\n".join( cur_transcript_lines ) + "\n" )
kept_lines += len( cur_transcript_lines )
for i, feature in enumerate( GFFReaderWrapper( open( input_name ), fix_strand=True ) ):
count = 0
for interval in feature.intervals:
if interval.feature == feature_name:
count += 1
if eval( '%s %s' % ( count, condition ) ):
# Keep feature.
for interval in feature.intervals:
out.write( "\t".join(interval.fields) + '\n' )
kept_features += 1
# Clean up.
out.close()
info_msg = "%i lines kept (%.2f%%) using condition %s. " % ( kept_lines, float(kept_lines)/i * 100.0, feature_name + condition )
info_msg = "%i of %i features kept (%.2f%%) using condition %s. " % \
( kept_features, i, float(kept_features)/i * 100.0, feature_name + condition )
if skipped_lines > 0:
info_msg += "Skipped %d blank/comment/invalid lines starting with line #%d." %( skipped_lines, first_skipped_line )
print info_msg