def __main__():
input_filename = sys.argv[1]
output_filename = sys.argv[2]
species = odict()
cur_size = 0
for components in iter_fasta_alignment(input_filename):
species_not_written = species.keys()
for component in components:
if component.species not in species:
species[component.species] = tempfile.TemporaryFile()
species[component.species].write("-" * cur_size)
species[component.species].write(component.text)
try:
species_not_written.remove(component.species)
except ValueError:
#this is a new species
pass
for spec in species_not_written:
species[spec].write("-" * len(components[0].text))
cur_size += len(components[0].text)
out = open(output_filename, 'wb')
for spec, f in species.iteritems():
f.seek(0)
out.write(">%s\n%s\n" % (spec, f.read()))
out.close()
def __main__():
input_filename = sys.argv[1]
output_filename = sys.argv[2]
species = odict()
cur_size = 0
for components in iter_fasta_alignment( input_filename ):
species_not_written = species.keys()
for component in components:
if component.species not in species:
species[component.species] = tempfile.TemporaryFile()
species[component.species].write( "-" * cur_size )
species[component.species].write( component.text )
try:
species_not_written.remove( component.species )
except ValueError:
#this is a new species
pass
for spec in species_not_written:
species[spec].write( "-" * len( components[0].text ) )
cur_size += len( components[0].text )
out = open( output_filename, 'wb' )
for spec, f in species.iteritems():
f.seek( 0 )
out.write( ">%s\n%s\n" % ( spec, f.read() ) )
out.close()
def parse_ndx( self, ndx ):
ndx_dict = odict()
index_group = ''
for line in ndx.split('\n'):
if line.startswith('['):
index_group = line.strip(' []\n\r')
ndx_dict[ index_group ] = []
elif line.strip():
ndx_dict[ index_group ] += line.split()
return ndx_dict
def parse_ndx(self, ndx):
ndx_dict = odict()
index_group = ''
for line in ndx.split('\n'):
if line.startswith('['):
index_group = line.strip(' []\n\r')
ndx_dict[index_group] = []
elif line.strip():
ndx_dict[index_group] += line.split()
return ndx_dict
def read_unordered_gtf( iterator, strict=False ):
"""
Returns GTF features found in an iterator. GTF lines need not be ordered
or clustered for reader to work. Reader returns GFFFeature objects sorted
by transcript_id, chrom, and start position.
"""
# -- Get function that generates line/feature key. --
get_transcript_id = lambda fields: parse_gff_attributes( fields[8] )[ 'transcript_id' ]
if strict:
# Strict GTF parsing uses transcript_id only to group lines into feature.
key_fn = get_transcript_id
else:
# Use lenient parsing where chromosome + transcript_id is the key. This allows
# transcripts with same ID on different chromosomes; this occurs in some popular
# datasources, such as RefGenes in UCSC.
key_fn = lambda fields: fields[0] + '_' + get_transcript_id( fields )
# Aggregate intervals by transcript_id and collect comments.
feature_intervals = odict()
comments = []
for count, line in enumerate( iterator ):
if line.startswith( '#' ):
comments.append( Comment( line ) )
continue
line_key = key_fn( line.split('\t') )
if line_key in feature_intervals:
feature = feature_intervals[ line_key ]
else:
feature = []
feature_intervals[ line_key ] = feature
feature.append( GFFInterval( None, line.split( '\t' ) ) )
# Create features.
chroms_features = {}
for count, intervals in enumerate( feature_intervals.values() ):
# Sort intervals by start position.
intervals.sort( lambda a,b: cmp( a.start, b.start ) )
feature = GFFFeature( None, intervals=intervals )
if feature.chrom not in chroms_features:
chroms_features[ feature.chrom ] = []
chroms_features[ feature.chrom ].append( feature )
# Sort features by chrom, start position.
chroms_features_sorted = []
for chrom_features in chroms_features.values():
chroms_features_sorted.append( chrom_features )
chroms_features_sorted.sort( lambda a,b: cmp( a[0].chrom, b[0].chrom ) )
for features in chroms_features_sorted:
features.sort( lambda a,b: cmp( a.start, b.start ) )
# Yield comments first, then features.
# FIXME: comments can appear anywhere in file, not just the beginning.
# Ideally, then comments would be associated with features and output
# just before feature/line.
for comment in comments:
yield comment
for chrom_features in chroms_features_sorted:
for feature in chrom_features:
yield feature
def read_unordered_gtf(iterator, strict=False):
"""
Returns GTF features found in an iterator. GTF lines need not be ordered
or clustered for reader to work. Reader returns GFFFeature objects sorted
by transcript_id, chrom, and start position.
"""
# -- Get function that generates line/feature key. --
get_transcript_id = lambda fields: parse_gff_attributes(fields[8])[
'transcript_id']
if strict:
# Strict GTF parsing uses transcript_id only to group lines into feature.
key_fn = get_transcript_id
else:
# Use lenient parsing where chromosome + transcript_id is the key. This allows
# transcripts with same ID on different chromosomes; this occurs in some popular
# datasources, such as RefGenes in UCSC.
key_fn = lambda fields: fields[0] + '_' + get_transcript_id(fields)
# Aggregate intervals by transcript_id and collect comments.
feature_intervals = odict()
comments = []
for count, line in enumerate(iterator):
if line.startswith('#'):
comments.append(Comment(line))
continue
line_key = key_fn(line.split('\t'))
if line_key in feature_intervals:
feature = feature_intervals[line_key]
else:
feature = []
feature_intervals[line_key] = feature
feature.append(GFFInterval(None, line.split('\t')))
# Create features.
chroms_features = {}
for count, intervals in enumerate(feature_intervals.values()):
# Sort intervals by start position.
intervals.sort(lambda a, b: cmp(a.start, b.start))
feature = GFFFeature(None, intervals=intervals)
if feature.chrom not in chroms_features:
chroms_features[feature.chrom] = []
chroms_features[feature.chrom].append(feature)
# Sort features by chrom, start position.
chroms_features_sorted = []
for chrom_features in chroms_features.values():
chroms_features_sorted.append(chrom_features)
chroms_features_sorted.sort(lambda a, b: cmp(a[0].chrom, b[0].chrom))
for features in chroms_features_sorted:
features.sort(lambda a, b: cmp(a.start, b.start))
# Yield comments first, then features.
# FIXME: comments can appear anywhere in file, not just the beginning.
# Ideally, then comments would be associated with features and output
# just before feature/line.
for comment in comments:
yield comment
for chrom_features in chroms_features_sorted:
for feature in chrom_features:
yield feature
