def main():
# Arguments
input_fname, out_fname = sys.argv[1:]
# Do conversion.
index = Indexes()
offset = 0
reader_wrapper = GFFReaderWrapper(fileinput.FileInput(input_fname), fix_strand=True)
for feature in list(reader_wrapper):
# Add feature; index expects BED coordinates.
if isinstance(feature, GenomicInterval):
convert_gff_coords_to_bed(feature)
index.add(feature.chrom, feature.start, feature.end, offset)
# Always increment offset, even if feature is not an interval and hence
# not included in the index.
offset += feature.raw_size
index.write(open(out_fname, "w"))
def main():
# Arguments
input_fname, out_fname = sys.argv[1:]
# Do conversion.
index = Indexes()
offset = 0
reader_wrapper = GFFReaderWrapper(fileinput.FileInput(input_fname),
fix_strand=True)
for feature in list(reader_wrapper):
# Add feature; index expects BED coordinates.
if isinstance(feature, GenomicInterval):
convert_gff_coords_to_bed(feature)
index.add(feature.chrom, feature.start, feature.end, offset)
# Always increment offset, even if feature is not an interval and hence
# not included in the index.
offset += feature.raw_size
index.write(open(out_fname, "wb"))
def main():
# Process arguments.
parser = optparse.OptionParser()
parser.add_option( '-F', '--format', dest="input_format" )
(options, args) = parser.parse_args()
in_fname, out_fname = args
input_format = options.input_format.lower()
# Create dict of name-location pairings.
name_loc_dict = {}
if input_format in [ 'gff', 'gtf' ]:
# GTF/GFF format
# Create reader.
if input_format == 'gff':
in_reader = GFFReaderWrapper( open( in_fname, 'r' ) )
else: # input_format == 'gtf'
in_reader = read_unordered_gtf( open( in_fname, 'r' ) )
for feature in in_reader:
if isinstance( feature, Comment ):
continue
for name in feature.attributes:
val = feature.attributes[ name ]
try:
float( val )
continue
except:
convert_gff_coords_to_bed( feature )
# Value is not a number, so it can be indexed.
if val not in name_loc_dict:
# Value is not in dictionary.
name_loc_dict[ val ] = {
'contig': feature.chrom,
'start': feature.start,
'end': feature.end
}
else:
# Value already in dictionary, so update dictionary.
loc = name_loc_dict[ val ]
if feature.start < loc[ 'start' ]:
loc[ 'start' ] = feature.start
if feature.end > loc[ 'end' ]:
loc[ 'end' ] = feature.end
elif input_format == 'bed':
# BED format.
for line in open( in_fname, 'r' ):
# Ignore track lines.
if line.startswith("track"):
continue
fields = line.split()
# Ignore lines with no feature name.
if len( fields ) < 4:
continue
# Process line
name_loc_dict[ fields[3] ] = {
'contig': fields[0],
'start': int( fields[1] ),
'end': int( fields[2] )
}
# Create sorted list of entries.
out = open( out_fname, 'w' )
max_len = 0
entries = []
for name in sorted( name_loc_dict.iterkeys() ):
loc = name_loc_dict[ name ]
entry = '%s\t%s\t%s' % ( name.lower(), name, '%s:%i-%i' % ( loc[ 'contig' ], loc[ 'start' ], loc[ 'end' ] ) )
if len( entry ) > max_len:
max_len = len( entry )
entries.append( entry )
# Write padded entries.
out.write( str( max_len + 1 ).ljust( max_len ) + '\n' )
for entry in entries:
out.write( entry.ljust( max_len ) + '\n' )
out.close()
twobitfile = None
file_iterator = open(input_filename)
if gff_format and interpret_features:
file_iterator = gff_util.GFFReaderWrapper(file_iterator,
fix_strand=False)
line_count = 1
for feature in file_iterator:
# Ignore comments, headers.
if isinstance(feature, (Header, Comment)):
line_count += 1
continue
name = ""
if gff_format and interpret_features:
# Processing features.
gff_util.convert_gff_coords_to_bed(feature)
chrom = feature.chrom
start = feature.start
end = feature.end
strand = feature.strand
else:
# Processing lines, either interval or GFF format.
line = feature.rstrip('\r\n')
if line and not line.startswith("#"):
fields = line.split('\t')
try:
chrom = fields[chrom_col]
start = int(fields[start_col])
end = int(fields[end_col])
if name_col:
name = fields[name_col]
def __main__():
#
# Parse options, args.
#
options, args = doc_optparse.parse(__doc__)
try:
if len(options.cols.split(',')) == 5:
# BED file
chrom_col, start_col, end_col, strand_col, name_col = parse_cols_arg(
options.cols)
else:
# gff file
chrom_col, start_col, end_col, strand_col = parse_cols_arg(
options.cols)
name_col = False
dbkey = options.dbkey
output_format = options.output_format
gff_format = options.gff
interpret_features = options.interpret_features
GALAXY_DATA_INDEX_DIR = options.GALAXY_DATA_INDEX_DIR
fasta_file = options.fasta
input_filename, output_filename = args
except:
doc_optparse.exception()
includes_strand_col = strand_col >= 0
strand = None
nibs = {}
#
# Set path to sequence data.
#
if fasta_file:
# Need to create 2bit file from fasta file.
try:
seq_path = tempfile.NamedTemporaryFile(dir=".").name
cmd = "faToTwoBit %s %s" % (fasta_file, seq_path)
tmp_name = tempfile.NamedTemporaryFile(dir=".").name
tmp_stderr = open(tmp_name, 'wb')
proc = subprocess.Popen(args=cmd,
shell=True,
stderr=tmp_stderr.fileno())
returncode = proc.wait()
tmp_stderr.close()
# Get stderr, allowing for case where it's very large.
tmp_stderr = open(tmp_name, 'rb')
stderr = ''
buffsize = 1048576
try:
while True:
stderr += tmp_stderr.read(buffsize)
if not stderr or len(stderr) % buffsize != 0:
break
except OverflowError:
pass
tmp_stderr.close()
# Error checking.
if returncode != 0:
raise Exception(stderr)
except Exception as e:
stop_err('Error running faToTwoBit. ' + str(e))
else:
seq_path = check_seq_file(dbkey, GALAXY_DATA_INDEX_DIR)
if not os.path.exists(seq_path):
# If this occurs, we need to fix the metadata validator.
stop_err(
"No sequences are available for '%s', request them by reporting this error."
% dbkey)
#
# Fetch sequences.
#
# Get feature's line(s).
def get_lines(feature):
if isinstance(feature, gff_util.GFFFeature):
return feature.lines()
else:
return [feature.rstrip('\r\n')]
skipped_lines = 0
first_invalid_line = 0
invalid_lines = []
fout = open(output_filename, "w")
warnings = []
warning = ''
twobitfile = None
file_iterator = open(input_filename)
if gff_format and interpret_features:
file_iterator = gff_util.GFFReaderWrapper(file_iterator,
fix_strand=False)
line_count = 1
for feature in file_iterator:
# Ignore comments, headers.
if isinstance(feature, (Header, Comment)):
line_count += 1
continue
name = ""
if gff_format and interpret_features:
# Processing features.
gff_util.convert_gff_coords_to_bed(feature)
chrom = feature.chrom
start = feature.start
end = feature.end
strand = feature.strand
else:
# Processing lines, either interval or GFF format.
line = feature.rstrip('\r\n')
if line and not line.startswith("#"):
fields = line.split('\t')
try:
chrom = fields[chrom_col]
start = int(fields[start_col])
end = int(fields[end_col])
if name_col:
name = fields[name_col]
if gff_format:
start, end = gff_util.convert_gff_coords_to_bed(
[start, end])
if includes_strand_col:
strand = fields[strand_col]
except:
warning = "Invalid chrom, start or end column values. "
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
if start > end:
warning = "Invalid interval, start '%d' > end '%d'. " % (
start, end)
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
if strand not in ['+', '-']:
strand = '+'
sequence = ''
else:
continue
# Open sequence file and get sequence for feature/interval.
if seq_path and os.path.exists("%s/%s.nib" % (seq_path, chrom)):
# TODO: improve support for GFF-nib interaction.
if chrom in nibs:
nib = nibs[chrom]
else:
nibs[chrom] = nib = bx.seq.nib.NibFile(
open("%s/%s.nib" % (seq_path, chrom)))
try:
sequence = nib.get(start, end - start)
except Exception as e:
warning = "Unable to fetch the sequence from '%d' to '%d' for build '%s'. " % (
start, end - start, dbkey)
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
elif seq_path and os.path.isfile(seq_path):
if not (twobitfile):
twobitfile = bx.seq.twobit.TwoBitFile(open(seq_path))
try:
if options.gff and interpret_features:
# Create sequence from intervals within a feature.
sequence = ''
for interval in feature.intervals:
sequence += twobitfile[
interval.chrom][interval.start:interval.end]
else:
sequence = twobitfile[chrom][start:end]
except:
warning = "Unable to fetch the sequence from '%d' to '%d' for chrom '%s'. " % (
start, end - start, chrom)
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
else:
warning = "Chromosome by name '%s' was not found for build '%s'. " % (
chrom, dbkey)
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
if sequence == '':
warning = "Chrom: '%s', start: '%s', end: '%s' is either invalid or not present in build '%s'. " % \
( chrom, start, end, dbkey )
warnings.append(warning)
if not invalid_lines:
invalid_lines = get_lines(feature)
first_invalid_line = line_count
skipped_lines += len(invalid_lines)
continue
if includes_strand_col and strand == "-":
sequence = reverse_complement(sequence)
if output_format == "fasta":
l = len(sequence)
c = 0
if gff_format:
start, end = gff_util.convert_bed_coords_to_gff([start, end])
fields = [dbkey, str(chrom), str(start), str(end), strand]
meta_data = "_".join(fields)
if name.strip():
fout.write(">%s %s\n" % (meta_data, name))
else:
fout.write(">%s\n" % meta_data)
while c < l:
b = min(c + 50, l)
fout.write("%s\n" % str(sequence[c:b]))
c = b
else: # output_format == "interval"
if gff_format and interpret_features:
# TODO: need better GFF Reader to capture all information needed
# to produce this line.
meta_data = "\t".join([
feature.chrom, "galaxy_extract_genomic_dna", "interval",
str(feature.start),
str(feature.end), feature.score, feature.strand, ".",
gff_util.gff_attributes_to_str(feature.attributes, "GTF")
])
else:
meta_data = "\t".join(fields)
if gff_format:
format_str = "%s seq \"%s\";\n"
else:
format_str = "%s\t%s\n"
fout.write(format_str % (meta_data, str(sequence)))
# Update line count.
if isinstance(feature, gff_util.GFFFeature):
line_count += len(feature.intervals)
else:
line_count += 1
fout.close()
if warnings:
warn_msg = "%d warnings, 1st is: " % len(warnings)
warn_msg += warnings[0]
print warn_msg
if skipped_lines:
# Error message includes up to the first 10 skipped lines.
print 'Skipped %d invalid lines, 1st is #%d, "%s"' % (
skipped_lines, first_invalid_line, '\n'.join(invalid_lines[:10]))
# Clean up temp file.
if fasta_file:
os.remove(seq_path)
os.remove(tmp_name)
def __main__():
#
# Parse options, args.
#
options, args = doc_optparse.parse( __doc__ )
try:
if len(options.cols.split(',')) == 5:
# BED file
chrom_col, start_col, end_col, strand_col, name_col = parse_cols_arg( options.cols )
else:
# gff file
chrom_col, start_col, end_col, strand_col = parse_cols_arg( options.cols )
name_col = False
dbkey = options.dbkey
output_format = options.output_format
gff_format = options.gff
interpret_features = options.interpret_features
GALAXY_DATA_INDEX_DIR = options.GALAXY_DATA_INDEX_DIR
fasta_file = options.fasta
input_filename, output_filename = args
except:
doc_optparse.exception()
includes_strand_col = strand_col >= 0
strand = None
nibs = {}
#
# Set path to sequence data.
#
if fasta_file:
# Need to create 2bit file from fasta file.
try:
seq_path = tempfile.NamedTemporaryFile( dir="." ).name
cmd = "faToTwoBit %s %s" % ( fasta_file, seq_path )
tmp_name = tempfile.NamedTemporaryFile( dir="." ).name
tmp_stderr = open( tmp_name, 'wb' )
proc = subprocess.Popen( args=cmd, shell=True, stderr=tmp_stderr.fileno() )
returncode = proc.wait()
tmp_stderr.close()
# Get stderr, allowing for case where it's very large.
tmp_stderr = open( tmp_name, 'rb' )
stderr = ''
buffsize = 1048576
try:
while True:
stderr += tmp_stderr.read( buffsize )
if not stderr or len( stderr ) % buffsize != 0:
break
except OverflowError:
pass
tmp_stderr.close()
# Error checking.
if returncode != 0:
raise Exception(stderr)
except Exception as e:
stop_err( 'Error running faToTwoBit. ' + str( e ) )
else:
seq_path = check_seq_file( dbkey, GALAXY_DATA_INDEX_DIR )
if not os.path.exists( seq_path ):
# If this occurs, we need to fix the metadata validator.
stop_err( "No sequences are available for '%s', request them by reporting this error." % dbkey )
#
# Fetch sequences.
#
# Get feature's line(s).
def get_lines( feature ):
if isinstance( feature, gff_util.GFFFeature ):
return feature.lines()
else:
return [ feature.rstrip( '\r\n' ) ]
skipped_lines = 0
first_invalid_line = 0
invalid_lines = []
fout = open( output_filename, "w" )
warnings = []
warning = ''
twobitfile = None
file_iterator = open( input_filename )
if gff_format and interpret_features:
file_iterator = gff_util.GFFReaderWrapper( file_iterator, fix_strand=False )
line_count = 1
for feature in file_iterator:
# Ignore comments, headers.
if isinstance( feature, ( Header, Comment ) ):
line_count += 1
continue
name = ""
if gff_format and interpret_features:
# Processing features.
gff_util.convert_gff_coords_to_bed( feature )
chrom = feature.chrom
start = feature.start
end = feature.end
strand = feature.strand
else:
# Processing lines, either interval or GFF format.
line = feature.rstrip( '\r\n' )
if line and not line.startswith( "#" ):
fields = line.split( '\t' )
try:
chrom = fields[chrom_col]
start = int( fields[start_col] )
end = int( fields[end_col] )
if name_col:
name = fields[name_col]
if gff_format:
start, end = gff_util.convert_gff_coords_to_bed( [start, end] )
if includes_strand_col:
strand = fields[strand_col]
except:
warning = "Invalid chrom, start or end column values. "
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
if start > end:
warning = "Invalid interval, start '%d' > end '%d'. " % ( start, end )
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
if strand not in ['+', '-']:
strand = '+'
sequence = ''
else:
continue
# Open sequence file and get sequence for feature/interval.
if seq_path and os.path.exists( "%s/%s.nib" % ( seq_path, chrom ) ):
# TODO: improve support for GFF-nib interaction.
if chrom in nibs:
nib = nibs[chrom]
else:
nibs[chrom] = nib = bx.seq.nib.NibFile( open( "%s/%s.nib" % ( seq_path, chrom ) ) )
try:
sequence = nib.get( start, end - start )
except Exception as e:
warning = "Unable to fetch the sequence from '%d' to '%d' for build '%s'. " % ( start, end - start, dbkey )
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
elif seq_path and os.path.isfile( seq_path ):
if not(twobitfile):
twobitfile = bx.seq.twobit.TwoBitFile( open( seq_path ) )
try:
if options.gff and interpret_features:
# Create sequence from intervals within a feature.
sequence = ''
for interval in feature.intervals:
sequence += twobitfile[interval.chrom][interval.start:interval.end]
else:
sequence = twobitfile[chrom][start:end]
except:
warning = "Unable to fetch the sequence from '%d' to '%d' for chrom '%s'. " % ( start, end - start, chrom )
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
else:
warning = "Chromosome by name '%s' was not found for build '%s'. " % ( chrom, dbkey )
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
if sequence == '':
warning = "Chrom: '%s', start: '%s', end: '%s' is either invalid or not present in build '%s'. " % \
( chrom, start, end, dbkey )
warnings.append( warning )
if not invalid_lines:
invalid_lines = get_lines( feature )
first_invalid_line = line_count
skipped_lines += len( invalid_lines )
continue
if includes_strand_col and strand == "-":
sequence = reverse_complement( sequence )
if output_format == "fasta":
l = len( sequence )
c = 0
if gff_format:
start, end = gff_util.convert_bed_coords_to_gff( [ start, end ] )
fields = [dbkey, str( chrom ), str( start ), str( end ), strand]
meta_data = "_".join( fields )
if name.strip():
fout.write( ">%s %s\n" % (meta_data, name) )
else:
fout.write( ">%s\n" % meta_data )
while c < l:
b = min( c + 50, l )
fout.write( "%s\n" % str( sequence[c:b] ) )
c = b
else: # output_format == "interval"
if gff_format and interpret_features:
# TODO: need better GFF Reader to capture all information needed
# to produce this line.
meta_data = "\t".join(
[feature.chrom, "galaxy_extract_genomic_dna", "interval",
str( feature.start ), str( feature.end ), feature.score, feature.strand,
".", gff_util.gff_attributes_to_str( feature.attributes, "GTF" ) ] )
else:
meta_data = "\t".join( fields )
if gff_format:
format_str = "%s seq \"%s\";\n"
else:
format_str = "%s\t%s\n"
fout.write( format_str % ( meta_data, str( sequence ) ) )
# Update line count.
if isinstance( feature, gff_util.GFFFeature ):
line_count += len( feature.intervals )
else:
line_count += 1
fout.close()
if warnings:
warn_msg = "%d warnings, 1st is: " % len( warnings )
warn_msg += warnings[0]
print(warn_msg)
if skipped_lines:
# Error message includes up to the first 10 skipped lines.
print('Skipped %d invalid lines, 1st is #%d, "%s"' % ( skipped_lines, first_invalid_line, '\n'.join( invalid_lines[:10] ) ))
# Clean up temp file.
if fasta_file:
os.remove( seq_path )
os.remove( tmp_name )
warning = ''
twobitfile = None
file_iterator = open( input_filename )
if gff_format and interpret_features:
file_iterator = gff_util.GFFReaderWrapper( file_iterator, fix_strand=False )
line_count = 1
for feature in file_iterator:
# Ignore comments, headers.
if isinstance( feature, ( Header, Comment ) ):
line_count += 1
continue
name = ""
if gff_format and interpret_features:
# Processing features.
gff_util.convert_gff_coords_to_bed( feature )
chrom = feature.chrom
start = feature.start
end = feature.end
strand = feature.strand
else:
# Processing lines, either interval or GFF format.
line = feature.rstrip( '\r\n' )
if line and not line.startswith( "#" ):
fields = line.split( '\t' )
try:
chrom = fields[chrom_col]
start = int( fields[start_col] )
end = int( fields[end_col] )
if name_col:
name = fields[name_col]
def main():
# Process arguments.
parser = optparse.OptionParser()
parser.add_option('-F', '--format', dest="input_format")
(options, args) = parser.parse_args()
in_fname, out_fname = args
input_format = options.input_format.lower()
# Create dict of name-location pairings.
name_loc_dict = {}
if input_format in ['gff', 'gtf']:
# GTF/GFF format
# Create reader.
if input_format == 'gff':
in_reader = GFFReaderWrapper(open(in_fname, 'r'))
else: #input_format == 'gtf'
in_reader = read_unordered_gtf(open(in_fname, 'r'))
for feature in in_reader:
if isinstance(feature, Comment):
continue
for name in feature.attributes:
val = feature.attributes[name]
try:
float(val)
continue
except:
convert_gff_coords_to_bed(feature)
# Value is not a number, so it can be indexed.
if val not in name_loc_dict:
# Value is not in dictionary.
name_loc_dict[val] = {
'contig': feature.chrom,
'start': feature.start,
'end': feature.end
}
else:
# Value already in dictionary, so update dictionary.
loc = name_loc_dict[val]
if feature.start < loc['start']:
loc['start'] = feature.start
if feature.end > loc['end']:
loc['end'] = feature.end
elif input_format == 'bed':
# BED format.
for line in open(in_fname, 'r'):
# Ignore track lines.
if line.startswith("track"):
continue
fields = line.split()
# Ignore lines with no feature name.
if len(fields) < 4:
continue
# Process line
name_loc_dict[fields[3]] = {
'contig': fields[0],
'start': int(fields[1]),
'end': int(fields[2])
}
# Create sorted list of entries.
out = open(out_fname, 'w')
max_len = 0
entries = []
for name in sorted(name_loc_dict.iterkeys()):
loc = name_loc_dict[name]
entry = '%s\t%s\t%s' % (name.lower(), name, '%s:%i-%i' %
(loc['contig'], loc['start'], loc['end']))
if len(entry) > max_len:
max_len = len(entry)
entries.append(entry)
# Write padded entries.
out.write(str(max_len + 1).ljust(max_len) + '\n')
for entry in entries:
out.write(entry.ljust(max_len) + '\n')
out.close()
