def pick_longest_rep(fasta_filename, gff_filename, group_filename, output_filename):
"""
For each group, select the representative record to be the longest
"""
fastad = LazyFastaReader(fasta_filename)
fout = FastaWriter(output_filename)
coords = {}
for line in open(gff_filename):
# ex: chr1 PacBio transcript 27567 29336 . - . gene_id "PB.1"; transcript_id "PB.1.1";
raw = line.strip().split("\t")
if raw[2] == "transcript":
tid = raw[-1].split("; ")[1].split()[1][1:-2]
coords[tid] = "{0}:{1}-{2}({3})".format(raw[0], raw[3], raw[4], raw[6])
for line in open(group_filename):
pb_id, members = line.strip().split("\t")
best_id = None
best_seq = None
max_len = 0
for x in members.split(","):
if len(fastad[x].sequence) >= max_len:
best_id = x
best_seq = fastad[x].sequence
max_len = len(fastad[x].sequence)
fout.writeRecord("{0}|{1}|{2}".format(pb_id, coords[pb_id], best_id), best_seq)
fout.close()
def main(argv):
desc = 'A tool to trim quiver results for contigs majority lowercase'
parser = argparse.ArgumentParser(description=desc)
parser.add_argument('inputFile', help='input sequence')
parser.add_argument('outputFile', help='output fasta')
parser.add_argument(
'--filt',
default=0.5,
dest='filt',
type=float,
help=
'proportion of lowercase bases a contig can have before being filtered out'
)
args = parser.parse_args()
writer = FastaWriter(args.outputFile)
for record in FastaReader(args.inputFile):
upper_output = []
upper_indx = []
lower = float(sum(1 for c in record.sequence if c.islower()))
pro = lower / float(len(record.sequence))
print pro
if pro < args.filt:
writer.writeRecord(record)
def separate_listed_sequences(fasta_file, good_values, good_output,
bad_output):
"""
Separate a fasta file into two based on a supplied value list
"""
with FastaWriter(good_output) as good_handle:
with FastaWriter(bad_output) as bad_handle:
for record in FastaReader(fasta_file):
name = get_base_sequence_name(record.name)
if name in good_values:
good_handle.writeRecord(record)
else:
bad_handle.writeRecord(record)
def separate_aligned_sequences(fasta_file, dictionary, good_values,
good_output, bad_output):
"""
Separate a fasta file into two based on a supplied dictionary and value list
"""
with FastaWriter(good_output) as good_handle:
with FastaWriter(bad_output) as bad_handle:
for record in FastaReader(fasta_file):
name = get_base_sequence_name(record.name)
value = dictionary.get(name, "Unmapped")
if value in good_values:
good_handle.writeRecord(record)
else:
bad_handle.writeRecord(record)
def _updateChimeraInfo(self,
suspicous_hits,
in_read_fn,
out_nc_fn,
out_c_fn,
primer_report_fn,
write_report_header=True):
"""
in_read_fn --- a fasta of full-length reads or a fasta of
non-full-length reads.
For each full-length read in in_read_fn FASTA file, detect whether
it is chimeric or not, and write its annotation to
primer_report_fn.
Return:
(num_nc, num_c, num_nc_bases, num_c_bases)
"""
logging.debug(
"Update chimera info for reads in {f} ".format(f=in_read_fn))
logging.debug(
"Write primer report to {rpt}".format(rpt=primer_report_fn))
num_nc, num_c, num_nc_bases, num_c_bases = 0, 0, 0, 0
with FastaReader(in_read_fn) as reader, \
FastaWriter(out_nc_fn) as writer, \
FastaWriter(out_c_fn) as writer_chimera, \
open(primer_report_fn, 'w') as reporter:
if write_report_header:
reporter.write(ReadAnnotation.header(delimiter=",") + "\n")
for r in reader:
# e.g. r.name="movie/zmw/0_100_CCS fiveend=1;threeend=100;"
readid = r.name.split()[0]
annotation = ReadAnnotation.fromString(
r.name, ignore_polyA=self.ignore_polyA)
if readid not in suspicous_hits: # Non-chimeric reads
# Primer of a primer-trimmed read can not be None.
# assert(annotation.primer is not None)
annotation.chimera = 0
num_nc += 1
num_nc_bases += len(r.sequence)
writer.writeRecord(annotation.toAnnotation(), r.sequence)
else: # chimeric reads
annotation.chimera = 1
num_c += 1
num_c_bases += len(r.sequence)
writer_chimera.writeRecord(annotation.toAnnotation(),
r.sequence)
reporter.write(annotation.toReportRecord(delimitor=",") + "\n")
return (num_nc, num_c, num_nc_bases, num_c_bases)
def writerProcess(outDir):
# makes output directories
if not os.path.exists(outDir):
os.makedirs(outDir)
fastOutDir = os.path.join(outDir, "Demultiplexed/")
if not os.path.exists(fastOutDir):
os.makedirs(fastOutDir)
# opens files
csvOut = open(os.path.join(outDir, "Report.csv"), "w")
csvOut.write("Name,Barcode,NumPasses,Coverage,AvgConfidence,MinConfidence,TrimFail,MappingFail\n")
writers = {}
for writecount in range(totalNumber):
result = resultQueue.get()
csvOut.write("%s,%s,%d,%d,%0.6f,%0.6f,%s,%s\n" % (
result.name, result.barcode, result.numPasses, result.coverage, result.predictedAccuracy,
result.minConfidence, result.trimFail, result.mappingFail))
if result.barcode not in writers:
if args.fastq:
writers[result.barcode] = FastqWriter(os.path.join(fastOutDir, result.barcode + ".fastq"))
else:
writers[result.barcode] = FastaWriter(os.path.join(fastOutDir, result.barcode + ".fasta"))
if not any((result.minNumPassesFail, result.mappingFail, result.trimFail, result.minCoverageFail,
result.minAvgConfidenceFail, result.minConfidenceFail)):
if args.fastq:
writers[result.barcode].writeRecord(result.name, result.seq, result.qual)
else:
writers[result.barcode].writeRecord(result.name, result.seq)
def _write_assigned_reads(input_fasta, assignments):
"""
Write out subreads to the appropriate file
"""
log.info("Separating subreads based on their amplicon assignments")
output_files = []
writers = {}
root_name = '.'.join(input_fasta.split('.')[:-1])
# Open up output writers for each group
for group in assignments:
output_file = "%s_%s.fasta" % (root_name, group)
output_files.append(output_file)
writers[group] = FastaWriter(output_file)
# Write each record to it's appropriate group(s)
for record in FastaReader(input_fasta):
name = record.name.split()[0]
for group in assignments:
if name in assignments[group]:
writers[group].writeRecord(record)
break
# Close all of the output writers
for group in writers:
writers[group].close()
return output_files
def open_writer(self):
if self.filetype == 'fasta':
output_file = '%s.trim.fasta' % self.prefix
self.writer = FastaWriter(output_file)
elif self.filetype == 'fastq':
output_file = '%s.trim.fastq' % self.prefix
self.writer = FastqWriter(output_file)
def add_writer(self, group):
if self.filetype == 'fasta':
output_file = '%s.g%s.fasta' % (self.prefix, group)
self.writers[group] = FastaWriter(output_file)
if self.filetype == 'fastq':
output_file = '%s.g%s.fastq' % (self.prefix, group)
self.writers[group] = FastqWriter(output_file)
def writeSequenceData(self, sequenceData):
outputFile = 'temp_%s.fasta' % self.counter
with FastaWriter(outputFile) as handle:
for record in sequenceData:
handle.writeRecord(record)
self.tempFiles.append(outputFile)
return outputFile
def pick_rep(fa_fq_filename,
gff_filename,
group_filename,
output_filename,
is_fq=False,
pick_least_err_instead=True):
"""
For each group, select the representative record
If is FASTA file (is_fa False) -- then always pick the longest one
If is FASTQ file (is_fq True) -- then
If pick_least_err_instead is True, pick the one w/ least number of expected base errors
Else, pick the longest one
"""
if is_fq:
fd = LazyFastqReader(fa_fq_filename)
fout = FastqWriter(output_filename)
else:
fd = LazyFastaReader(fa_fq_filename)
fout = FastaWriter(output_filename)
coords = {}
for line in open(gff_filename):
# ex: chr1 PacBio transcript 27567 29336 . - . gene_id "PB.1"; transcript_id "PB.1.1";
raw = line.strip().split('\t')
if raw[2] == 'transcript':
tid = raw[-1].split('; ')[1].split()[1][1:-2]
coords[tid] = "{0}:{1}-{2}({3})".format(raw[0], raw[3], raw[4],
raw[6])
for line in open(group_filename):
pb_id, members = line.strip().split('\t')
print >> sys.stderr, "Picking representative sequence for", pb_id
best_id = None
best_seq = None
best_qual = None
best_err = 9999999
err = 9999999
max_len = 0
for x in members.split(','):
if is_fq and pick_least_err_instead:
err = sum(i**-(i / 10.) for i in fd[x].quality)
if (is_fq and pick_least_err_instead and err < best_err) or (
(not is_fq or not pick_least_err_instead)
and len(fd[x].sequence) >= max_len):
best_id = x
best_seq = fd[x].sequence
if is_fq:
best_qual = fd[x].quality
best_err = err
max_len = len(fd[x].sequence)
_id_ = "{0}|{1}|{2}".format(pb_id, coords[pb_id], best_id)
_seq_ = best_seq
if is_fq:
fout.writeRecord(_id_, _seq_, best_qual)
else:
fout.writeRecord(_id_, _seq_)
fout.close()
def write_references(reference_file, references):
for i, ref in enumerate(references):
for record in FastaReader(reference_file):
name = record.name.split()[0]
if name == ref:
filename = 'reference_%s.fasta' % (i + 1)
with FastaWriter(filename) as writer:
writer.writeRecord(record)
def subset_references(reference_file, reference_names):
output = 'references.fasta'
with FastaWriter(output) as writer:
for record in FastaReader(reference_file):
name = record.name.split()[0]
if name in reference_names:
writer.writeRecord(record)
return output
def get_temp_fasta(record):
"""
Create a temporary Fasta file for Blasr/HMMsearch/etc
"""
temp_record = get_temp_fasta_record(record)
temp_fasta = NamedTemporaryFile(suffix='.fasta')
with FastaWriter(temp_fasta.name) as handle:
handle.writeRecord(temp_record)
return temp_fasta
def combine_fasta(fasta_files, destination):
with FastaWriter(destination) as handle:
for fasta in fasta_files:
try:
for record in FastaReader(fasta):
handle.writeRecord(record)
except:
log.warn('Could not open "%s" as Fasta' % fasta)
check_output_file(destination)
def write_fasta(records, output_file):
"""
Write a FastaRecord, or a list of FastaRecords, out to file
"""
with FastaWriter(output_file) as handle:
for record in records:
assert isinstance(record, FastaRecord)
handle.writeRecord(record)
check_output_file(output_file)
return output_file
def extract_subreads(input_file,
output_file,
min_length,
max_length,
min_score,
min_snr,
max_count,
white_list=None):
"""
Extract, filter and subset subreads from Bas/Bax/Fofn Files
"""
log.info('Extracting subreads from %s' % os.path.basename(input_file))
log.debug('\tMinimum Length:\t%s' % min_length)
log.debug('\tMaximum Length:\t%s' % max_length)
log.debug('\tMinimum Score:\t%s' % min_score)
log.debug('\tMinimum SNR:\t%s' % min_snr)
log.debug('\tMax Count:\t%s' % max_count)
log.debug('\tWhitelisted ZMWs:\t%s' % white_list)
if white_list:
white_list = set(_parse_white_list(white_list))
output_prefix = os.path.dirname(output_file)
output_file_list = []
subread_count = 0
for i, filename in enumerate(_iterate_input_files(input_file)):
curr_output = os.path.join(output_prefix,
'subreads_%s.fasta' % (i + 1))
if filename.endswith('.bas.h5') or filename.endswith('bax.h5'):
subreads = _extract_from_bash5(filename, min_length, max_length,
min_score, min_snr, white_list)
elif filename.endswith('.fa') or filename.endswith('.fasta'):
subreads = _extract_from_fasta(filename, min_length, max_length)
with FastaWriter(curr_output) as writer:
for record in subreads:
writer.writeRecord(record)
subread_count += len(subreads)
output_file_list.append(curr_output)
log.info("Extracted %s subreads from %s files" % (subread_count, i + 1))
log.info("Writing FOFN of subread files")
with open(output_file, 'w') as handle:
for filename in output_file_list:
handle.write(filename + '\n')
# TODO: Fix MaxCount function
#if max_count:
# subreads = _subset_subreads( subreads, max_count )
log.info("Finished extracting subreads")
return output_file
def make_current_fasta(icec_obj, flnc_filename, root_dir):
"""
current fasta will consists of all ids
however --- if this was a already finished run and we are adding more input,
then newids is empty, in this case we set newids = everything that
has no affiliation or more than one affiliated cluster in d
"""
with FastaWriter(current_fasta(root_dir)) as f:
for r in FastaReader(flnc_filename):
f.writeRecord(r)
def _open_output_handle(output_file, output_type):
"""
Open an appropriate output handle to record the exon sequences
"""
if output_type == 'fasta':
return FastaWriter(output_file)
elif output_type == 'fastq':
return FastqWriter(output_file)
msg = 'Output type must be Fasta or Fastq'
log.error(msg)
raise TypeError(msg)
def outputReferenceFasta(self, reference, count):
print "Creating reference sequence for Cluster #%s" % count
referenceFile = 'cluster%s_ref.fasta' % count
reference_desc = 'cluster{0}_reference\t{1}'.format(
count, reference.name)
if os.path.exists(referenceFile):
return referenceFile
with FastaWriter(referenceFile) as handle:
referenceFasta = FastaRecord(reference_desc, reference.sequence)
handle.writeRecord(referenceFasta)
return referenceFile
def outputClusterFasta(self, reads, count):
fastaFile = 'cluster%s.fasta' % count
if os.path.exists(fastaFile):
return fastaFile
# Rename the "Reference" sequence to the cluster
with FastaWriter(fastaFile) as handle:
for fastqRecord in reads:
fastaRecord = FastaRecord(fastqRecord.name,
fastqRecord.sequence)
handle.writeRecord(fastaRecord)
return fastaFile
def convert_to_dazz_fasta(self):
"""
Convert input fasta/fastq file to daligner-compatibe fasta with ids:
<prefix>/<index>/0_<seqlen>
Also write out mappings to pickle
"""
i = 1
reader = FastaReader(self.input_filename) if self.filetype == "fasta" else FastqReader(self.input_filename)
f = FastaWriter(self.dazz_filename)
for r in reader:
f.writeRecord("{p}/{i}/0_{len}".format(p=self.dazz_movie_name, i=i, len=len(r.sequence)), r.sequence)
self.dazz_mapping[i] = r.id
i += 1
f.close()
with open(self.dazz_filename + ".pickle", "w") as f:
dump(self.dazz_mapping, f)
def pick_longest_rep(fasta_filename, gff_filename, group_filename,
output_filename):
"""
For each group, select the representative record to be the longest
"""
fastad = LazyFastaReader(fasta_filename)
fout = FastaWriter(output_filename)
coords = {}
for line in open(gff_filename):
# ex: chr1 PacBio transcript 27567 29336 . - . gene_id "PB.1"; transcript_id "PB.1.1";
raw = line.strip().split('\t')
if raw[2] == 'transcript':
tid = raw[-1].split('; ')[1].split()[1][1:-2]
coords[tid] = "{0}:{1}-{2}({3})".format(raw[0], raw[3], raw[4],
raw[6])
for line in open(group_filename):
pb_id, members = line.strip().split('\t')
best_id = None
best_seq = None
max_len = 0
for x in members.split(','):
if len(fastad[x].sequence) >= max_len:
best_id = x
best_seq = fastad[x].sequence
max_len = len(fastad[x].sequence)
fout.writeRecord("{0}|{1}|{2}".format(pb_id, coords[pb_id], best_id),
best_seq)
fout.close()
def runBlasr(cls, fastqRecord, alignedRecord):
# Write the query and reference records to file
tempId = str(int(random() * 10000000))
tempRef = 'temp_ref_%s.fasta' % tempId
with FastaWriter(tempRef) as handle:
fastaRecord = cls.convertFastqToFasta(fastqRecord)
handle.writeRecord(fastaRecord)
tempQuery = 'temp_query_%s.fasta' % tempId
with FastaWriter(tempQuery) as handle:
handle.writeRecord(alignedRecord)
# Create and run the command-line
tempOut = 'temp_%s.m1' % tempId
cline = 'blasr %s %s -m 1 -bestn 1 -out %s' % (tempQuery, tempRef,
tempOut)
p = subprocess.Popen(cline.split())
stdout, stderr = p.communicate()
# Parse and return the best hit and remove temp files
bestHit = cls.readBestBlasrHit(tempOut)
os.remove(tempRef)
os.remove(tempQuery)
os.remove(tempOut)
return bestHit
def combine_fasta(sequence_files, output_file):
"""
Combine a series of sequence files into one Fasta
"""
with FastaWriter(output_file) as handle:
for filename in sequence_files:
try:
for record in FastaReader(filename):
handle.writeRecord(record)
except:
log.warn('Could not open "%s" as Fasta' % fasta)
check_output_file(output_file)
return output_file
def output_final_sequences(self, finalSequenceList):
outputFile = self.process_setup(finalSequenceList,
'SequenceWriter',
suffix='fasta')
if self.output_files_exist(output_file=outputFile):
return outputFile
with FastaWriter(outputFile) as writer:
with open(finalSequenceList) as handle:
for line in handle:
sequenceFile = line.strip()
copy_fasta_sequences(sequenceFile, writer)
self.process_cleanup(output_file=outputFile)
return outputFile
def write_fasta(fasta_records, output_file):
"""
Write a FastaRecord, or list of records, out to file
"""
with FastaWriter(output_file) as handle:
if isinstance(fasta_records, FastaRecord):
handle.writeRecord(fasta_records)
elif isinstance(fasta_records, list):
for record in fasta_records:
handle.writeRecord(record)
else:
msg = "Input Record(s) type not recognized"
log.error(msg)
raise TypeError(msg)
check_output_file(output_file)
def separate_sequences(fasta_file, dictionary, prefix=''):
"""
Separate a fasta file into multiple groups based on some dict
"""
file_handles = {}
for record in FastaReader(fasta_file):
name = get_base_sequence_name(record.name)
group = dictionary.get(name, "Unmapped")
group_file = prefix + '_' + group + '.fasta'
try:
file_handles[group_file].writeRecord(record)
except KeyError:
file_handles[group_file] = FastaWriter(group_file)
file_handles[group_file].writeRecord(record)
return closed_file_handles(file_handles)
def _updateChimeraInfo(self, suspicous_hits, in_read_fn, out_flnc_fn,
out_flc_fn, primer_report_fl_fn):
"""
in_read_fn --- a fasta of full-length reads
For each full-length read in in_read_fn FASTA file, detect whether
it is chimeric or not, and write its annotation to
primer_report_fl_fn.
"""
logging.info("Update chimera info to reads annotations " +
"in the output FASTA file and the primer report.")
with FastaReader(in_read_fn) as reader, \
FastaWriter(out_flnc_fn) as writer, \
FastaWriter(out_flc_fn) as writer_chimera, \
open(primer_report_fl_fn, 'w') as reporter:
reporter.write("\t".join(ReadAnnotation.fieldsNames()) + "\n")
for r in reader:
# e.g. r.name="movie/zmw/0_100_CCS fiveend=1;threeend=100;"
readid = r.name.split()[0]
annotation = ReadAnnotation.fromString(
r.name, ignore_polyA=self.ignore_polyA)
if readid not in suspicous_hits: # Non-chimeric reads
# Primer of a primer-trimmed read can not be None.
# assert(annotation.primer is not None)
annotation.chimera = 0
assert (annotation.isFullLength)
self.summary.num_flnc += 1
self.summary.num_flnc_bases += len(r.sequence)
writer.writeRecord(annotation.toAnnotation(), r.sequence)
else: # chimeric reads
annotation.chimera = 1
self.summary.num_flc += 1
writer_chimera.writeRecord(annotation.toAnnotation(),
r.sequence)
reporter.write(annotation.toReportRecord() + "\n")
def extract_exons(afa_file, info_file):
locus = afa_file.split('_')[0]
output_fofn = '%s_exons.fofn' % locus
records = list(FastaReader(afa_file))
regions = list(_parse_info_file(info_file))
with open(output_fofn, 'w') as fofn_handle:
for exon in _select_exons(regions):
output_file = _get_output_file(locus, exon)
with FastaWriter(output_file) as output:
for record in _extract_fasta_region(records, exon):
if len(set(record.sequence)) == 1:
continue
output.writeRecord(record)
fofn_handle.write(os.path.abspath(output_file) + '\n')
def rename_fasta( input_file, output_file, name_key ):
"""
Rename a single Fasta of subreads
"""
renaming_dict = read_dict_file( name_key )
with FastaWriter( output_file ) as writer:
for record in FastaReader( input_file ):
old_name = record.name.split()[0]
try:
new_name = renaming_dict[old_name]
except KeyError:
msg = "Sequence name not found!"
log.error( msg )
raise KeyError( msg )
new_record = FastaRecord( new_name, record.sequence )
writer.writeRecord( new_record )
check_output_file( output_file )
return output_file
def extract_subreads(input_file,
output_file,
min_length,
max_length,
min_score,
min_snr,
max_count,
white_list=None):
"""
Extract, filter and subset subreads from Bas/Bax/Fofn Files
"""
log.info('Extracting subreads from %s' % os.path.basename(input_file))
log.debug('\tMinimum Length:\t%s' % min_length)
log.debug('\tMaximum Length:\t%s' % max_length)
log.debug('\tMinimum Score:\t%s' % min_score)
log.debug('\tMinimum SNR:\t%s' % min_snr)
log.debug('\tMax Count:\t%s' % max_count)
log.debug('\tWhitelisted ZMWs:\t%s' % white_list)
if white_list:
white_list = set(_parse_white_list(white_list))
subreads = []
for i, filename in enumerate(_iterate_input_files(input_file)):
if filename.endswith('.bas.h5') or filename.endswith('bax.h5'):
subreads += _extract_from_bash5(filename, min_length, max_length,
min_score, min_snr, white_list)
elif filename.endswith('.fa') or filename.endswith('.fasta'):
subreads += _extract_from_fasta(filename, min_length, max_length)
log.info("Extracted %s subreads from %s files" % (len(subreads), i + 1))
if max_count:
subreads = _subset_subreads(subreads, max_count)
with FastaWriter(output_file) as writer:
for record in subreads:
writer.writeRecord(record)
log.info("Finished extracting subreads")
return output_file
