def write_to_genbank(self, filename: str = None, directory: str = None, record: SeqRecord = None) -> None:
""" Writes a genbank file containing only the information contained
within the Region.
"""
if not filename:
filename = "%s.region%03d.gbk" % (self.parent_record.id, self.get_region_number())
if directory:
filename = os.path.join(directory, filename)
if record is None:
record = self.parent_record.to_biopython()
assert isinstance(record, SeqRecord)
with warnings.catch_warnings():
warnings.simplefilter("ignore")
cluster_record = record[self.location.start:self.location.end]
cluster_record.annotations["date"] = record.annotations.get("date", '')
cluster_record.annotations["source"] = record.annotations.get("source", '')
cluster_record.annotations["organism"] = record.annotations.get("organism", '')
cluster_record.annotations["taxonomy"] = record.annotations.get("taxonomy", [])
cluster_record.annotations["data_file_division"] = record.annotations.get("data_file_division", 'UNK')
cluster_record.annotations["comment"] = record.annotations.get("comment", '')
# update the antiSMASH annotation to include some cluster details
comment_end_marker = "##antiSMASH-Data-END"
cluster_comment = ("NOTE: This is a single cluster extracted from a larger record!\n"
"Orig. start :: {start}\n"
"Orig. end :: {end}\n"
"{end_marker}").format(start=self.location.start,
end=self.location.end,
end_marker=comment_end_marker)
original = cluster_record.annotations["comment"]
cluster_record.annotations["comment"] = original.replace(comment_end_marker, cluster_comment)
# our cut-out clusters are always linear
cluster_record.annotations["topology"] = "linear"
# renumber clusters, superclusters and regions to reflect changes
first_supercluster = min(sc.get_supercluster_number() for sc in self.superclusters)
first_cluster = min(cluster.get_cluster_number() for cluster in self.get_unique_clusters())
first_subregion = min(sub.get_subregion_number() for sub in self.subregions) if self.subregions else 0
for feature in cluster_record.features:
if feature.type == "region":
supers = feature.qualifiers.get("candidate_cluster_numbers")
if not supers:
continue
feature.qualifiers["candidate_cluster_numbers"] = [str(int(num) - first_supercluster) for num in supers]
elif feature.type == SuperCluster.FEATURE_TYPE:
new = str(int(feature.qualifiers["candidate_cluster_number"][0]) - first_supercluster)
feature.qualifiers["candidate_cluster_number"] = [new]
new_clusters = [str(int(num) - first_cluster) for num in feature.qualifiers["protoclusters"]]
feature.qualifiers["protoclusters"] = new_clusters
elif feature.type in ["protocluster", "protocluster_core"]:
new = str(int(feature.qualifiers["protocluster_number"][0]) - first_cluster)
feature.qualifiers["cluster_number"] = [new]
elif feature.type == "subregion":
new = str(int(feature.qualifiers["subregion_number"][0]) - first_subregion)
feature.qualifiers["subregion_number"] = [new]
seqio.write([cluster_record], filename, 'genbank')
def run_prodigal(record: Record, options: ConfigType) -> None:
""" Run progidal to annotate prokaryotic sequences
"""
if "basedir" in options.get('prodigal', ''):
basedir = options.prodigal.basedir
else:
basedir = ""
with TemporaryDirectory(change=True):
name = record.id.lstrip('-')
if not name:
name = "unknown"
fasta_file = '%s.fasta' % name
result_file = '%s.predict' % name
with open(fasta_file, 'w') as handle:
seqio.write([record.to_biopython()], handle, 'fasta')
# run prodigal
prodigal = [path.join(basedir, 'prodigal')]
prodigal.extend(['-i', fasta_file, '-f', 'sco', '-o', result_file])
if options.genefinding_tool == "prodigal-m" or len(record.seq) < 20000:
prodigal.extend(['-p', 'meta'])
err = execute(prodigal).stderr
if err.find('Error') > -1:
logging.error("Failed to run prodigal: %r", err)
raise RuntimeError("prodigal error: %s" % err)
found = 0
for line in open(result_file, 'r'):
# skip first line
if not line.startswith('>'):
continue
name, start_chunk, end_chunk, prodigal_strand = line[1:].rstrip(
).split("_")
try:
start = int(start_chunk)
end = int(end_chunk)
if prodigal_strand == "+":
strand = 1
else:
strand = -1
except ValueError:
logging.error('Malformatted prodigal output line %r',
line.rstrip())
continue
if start > end:
strand = -1
start, end = end, start
loc = FeatureLocation(start - 1, end, strand=strand)
translation = record.get_aa_translation_from_location(loc)
feature = CDSFeature(loc,
locus_tag='ctg%s_%s' %
(record.record_index, name),
translation=translation,
translation_table=record.transl_table)
record.add_cds_feature(feature)
found += 1
logging.debug("prodigal found %d CDS features", found)
def write(seq_records, options):
basename = seq_records[0].id
output_name = path.join(options.outputfoldername,
"%s.final.embl" % basename)
logging.debug("Writing seq_records to %r" % output_name)
if options.input_type == 'nucl':
seqio.write(seq_records, output_name, 'embl')
def write_to_genbank(self, filename=None, directory=None, record=None):
""" Writes a genbank file containing only the information contained
within the Cluster.
"""
if not filename:
filename = "%s.cluster%03d.gbk" % (self.parent_record.id,
self.get_cluster_number())
if directory:
filename = os.path.join(directory, filename)
if record is None:
record = self.parent_record.to_biopython()
assert isinstance(record, SeqRecord)
with warnings.catch_warnings():
warnings.simplefilter("ignore")
cluster_record = record[self.location.start:self.location.end]
cluster_record.annotations["date"] = record.annotations.get("date", '')
cluster_record.annotations["source"] = record.annotations.get(
"source", '')
cluster_record.annotations["organism"] = record.annotations.get(
"organism", '')
cluster_record.annotations["taxonomy"] = record.annotations.get(
"taxonomy", [])
cluster_record.annotations[
"data_file_division"] = record.annotations.get(
"data_file_division", 'UNK')
# our cut-out clusters are always linear
cluster_record.annotations["topology"] = "linear"
seqio.write([cluster_record], filename, 'genbank')
def write_search_fasta(record: Record) -> str:
""" Constructs a FASTA representation of a record and writes it to a
file in the current directory.
Returns:
the name of the file created
"""
filename = "{}.fasta".format(record.id)
with open(filename, 'w') as handle:
seqio.write([record.to_biopython()], handle, 'fasta')
return filename
def write(seq_records, options):
basename = seq_records[0].id
if options.input_type == 'nucl':
output_name = path.join(options.outputfoldername, "%s.final.gbk" % basename)
logging.debug("Writing %s seq_records to %r" % (len(seq_records), output_name))
seqio.write(seq_records, output_name, 'genbank')
i=1
for rec in seq_records:
# For compatibility with the database importer, we have to check whether we are dealing
# with a seq_record obtained from a file (then its class will be SeqRecord) or from a#
# database (then its class will be DBSeqRecord)
#
# running the cluster extraction on a DBSeqRecord will throw an exception, as splitting the object is not supported
if rec.__class__.__name__ == 'SeqRecord':
for cluster in utils.get_cluster_features(rec):
with warnings.catch_warnings():
warnings.simplefilter("ignore")
cluster_rec = rec[cluster.location.start:cluster.location.end]
output_name = path.join(options.outputfoldername,
"%s.cluster%03d.gbk" % (basename, i))
seqio.write([cluster_rec], output_name, 'genbank')
i += 1
else:
seq_records = seq_record_convert_nucl_to_prot(seq_records, options)
output_name = path.join(options.outputfoldername, "%s.final.gp" % basename)
logging.debug("Writing seq_records to %r" % output_name)
seqio.write(seq_records, output_name, 'genbank')
def write(seq_records, options):
basename = seq_records[0].id
if options.input_type == 'nucl':
output_name = path.join(options.outputfoldername,
"%s.final.gbk" % basename)
for rec in seq_records:
for cluster in utils.get_cluster_features(rec):
with warnings.catch_warnings():
warnings.simplefilter("ignore")
cluster_rec = rec[cluster.location.start:cluster.location.
end]
cluster_rec.annotations["date"] = rec.annotations.get(
"date", '')
cluster_rec.annotations["source"] = rec.annotations.get(
"source", '')
cluster_rec.annotations["organism"] = rec.annotations.get(
"organism", '')
cluster_rec.annotations["taxonomy"] = rec.annotations.get(
"taxonomy", [])
cluster_rec.annotations[
"data_file_division"] = rec.annotations.get(
"data_file_division", 'UNK')
# our cut-out clusters are always linear
cluster_rec.annotations["topology"] = "linear"
cluster_name = path.join(
options.outputfoldername, "%s.cluster%03d.gbk" %
(basename, utils.get_cluster_number(cluster)))
seqio.write([cluster_rec], cluster_name, 'genbank')
else:
seq_records = seq_record_convert_nucl_to_prot(seq_records, options)
output_name = path.join(options.outputfoldername,
"%s.final.gp" % basename)
logging.debug("Writing seq_records to %r" % output_name)
seqio.write(seq_records, output_name, 'genbank')
def run_glimmer(seq_record, options):
"Run glimmer3 to annotate prokaryotic sequences"
basedir = utils.get_genefinding_basedir(options)
with TemporaryDirectory(change=True):
utils.fix_record_name_id(seq_record, options)
name = seq_record.id
while len(name) > 0 and name[0] == '-':
name = name[1:]
if name == "":
name = "unknown"
fasta_file = '%s.fasta' % name
longorfs_file = '%s.longorfs' % name
icm_file = '%s.icm' % name
result_file = '%s.predict' % name
# run long-orfs
with open(fasta_file, 'w') as handle:
seqio.write([seq_record], handle, 'fasta')
long_orfs = [path.join(basedir, 'long-orfs')]
long_orfs.extend([
'-l', '-n', '-t', '1.15', '--trans_table', '11', fasta_file,
longorfs_file
])
out, err, _ = execute(long_orfs)
if err.find('ERROR') > -1:
logging.error("Locating long orfs failed: %r" % err)
return
# run extract
extract = [
path.join(basedir, 'extract'), '-t', fasta_file, longorfs_file
]
out, err, retcode = execute(extract)
if out == '':
logging.error("Failed to extract genes from model, aborting: %r" %
err)
return
build_icm = [path.join(basedir, 'build-icm'), '-r', icm_file]
out, err, retcode = execute(build_icm, input=out)
if err != '':
logging.error("Failed to build gene model: %r" % err)
return
# run glimmer3
glimmer = [path.join(basedir, 'glimmer3')]
glimmer.extend([
'-l', '-o', '50', '-g', '90', '-q', '3000', '-t', '30',
'--trans_table', '11', fasta_file, icm_file, name
])
out, err, retcode = execute(glimmer)
if err.find('ERROR') > -1:
logging.error("Failed to run glimmer3: %r" % err)
return
for line in open(result_file, 'r'):
# skip first line
if line.startswith('>'):
continue
name, start, end, strand, score = line.split()
try:
start = int(start)
end = int(end)
strand = int(strand)
except ValueError:
logging.error('Malformatted glimmer output line %r' %
line.rstrip())
if start > end:
bpy_strand = -1
tmp = start
start = end
end = tmp
else:
bpy_strand = 1
loc = FeatureLocation(start - 1, end, strand=bpy_strand)
feature = SeqFeature(location=loc,
id=name,
type="CDS",
qualifiers={
'locus_tag':
['ctg%s_%s' % (options.record_idx, name)],
'note': ['Glimmer score: %s' % score]
})
seq_record.features.append(feature)
def run_prodigal(seq_record, options):
"Run progidal to annotate prokaryotic sequences"
if "prodigal" in options:
if "basedir" in options.prodigal:
basedir = options.prodigal.basedir
else:
basedir = ""
with TemporaryDirectory(change=True):
utils.fix_record_name_id(seq_record, options)
name = seq_record.id
while len(name) > 0 and name[0] == '-':
name = name[1:]
if name == "":
name = "unknown"
fasta_file = '%s.fasta' % name
result_file = '%s.predict' % name
with open(fasta_file, 'w') as handle:
seqio.write([seq_record], handle, 'fasta')
# run prodigal
prodigal = [path.join(basedir, 'prodigal')]
prodigal.extend(['-i', fasta_file, '-f', 'sco', '-o', result_file])
if options.genefinding == "prodigal-m" or len(seq_record.seq) < 20000:
prodigal.extend(['-p', 'meta'])
err = execute(prodigal)[1]
if err.find('Error') > -1:
logging.error("Failed to run prodigal: %r" % err)
return
for line in open(result_file, 'r'):
# skip first line
if not line.startswith('>'):
continue
name, start, end, prodigalStrand = line[1:].rstrip().split("_")
try:
start = int(start)
end = int(end)
if prodigalStrand == "+":
strand = 1
else:
strand = -1
except ValueError:
logging.error('Malformatted prodigal output line %r' %
line.rstrip())
continue
if start > end:
strand = -1
tmp = start
start = end
end = tmp
loc = FeatureLocation(start - 1, end, strand=strand)
feature = SeqFeature(location=loc,
id=name,
type="CDS",
qualifiers={
'locus_tag':
['ctg%s_%s' % (options.record_idx, name)]
})
seq_record.features.append(feature)
def test_write_calls_biopython(self):
"Test writing Bio.SeqIO records"
mock("Bio.SeqIO.write", tracker=self.tt, returns=[])
expected_trace = " Called Bio.SeqIO.write(['fake'], DummyHandle('test.gbk'), 'genbank')"
seqio.write(['fake'], self.handle, "genbank")
assert_same_trace(self.tt, expected_trace)
def test_write_calls_biopython(self):
"Test writing Bio.SeqIO records"
mock("Bio.SeqIO.write", tracker=self.tt, returns=[])
expected_trace = " Called Bio.SeqIO.write(['fake'], DummyHandle('test.gbk'), 'genbank')"
seqio.write(['fake'], self.handle, "genbank")
assert_same_trace(self.tt, expected_trace)
def run_glimmerhmm(seq_record, options):
basedir = utils.get_genefinding_basedir(options)
with TemporaryDirectory(change=True):
#Write FASTA file and run GlimmerHMM
utils.fix_record_name_id(seq_record, options)
name = seq_record.id
while len(name) > 0 and name[0] == '-':
name = name[1:]
if name == "":
name = "unknown"
fasta_file = '%s.fasta' % name
result_file = '%s.predict' % name
with open(fasta_file, 'w') as handle:
seqio.write([seq_record], handle, 'fasta')
glimmerhmm = ['glimmerhmm']
glimmerhmm.extend([
fasta_file,
utils.get_full_path(__file__,
"train_%s" % options.glimmerhmm_train_folder),
"-g"
])
out, err, retcode = execute(glimmerhmm)
if err.find('ERROR') > -1:
logging.error("Failed to run GlimmerHMM: %r" % err)
return
#Parse GlimmerHMM predictions
resultstext = out
if "CDS" not in resultstext:
logging.error("GlimmerHMM gene prediction failed: no genes found.")
resultstext = resultstext.replace("\r", " ")
lines = resultstext.split("\n")
lines = lines[2:-1]
orfnames = []
positions = []
strands = []
x = 0
orfnr = 0
starts = []
ends = []
for line in lines:
columns = line.split("\t")
if len(columns) > 1:
if x == 0:
if columns[6] == "+":
bpy_strand = 1
else:
bpy_strand = -1
if "mRNA" not in line:
starts.append(int(columns[3]))
ends.append(int(columns[4]))
elif x == (len(lines) - 1) or "mRNA" in lines[x + 1]:
if columns[6] == "+":
bpy_strand = 1
else:
bpy_strand = -1
strands.append(bpy_strand)
starts.append(int(columns[3]))
ends.append(int(columns[4]))
orfnames.append("orf" + (5 - orfnr) * "0" + str(orfnr))
orfnr += 1
if len(starts) == 1:
if starts[0] == 0:
starts[0] = 1
if ends[0] == 0:
ends[0] = 1
positions.append([[starts[0] - 1, ends[0]]])
else:
pos = []
if bpy_strand == -1:
starts.reverse()
ends.reverse()
for i in starts:
if i == 0:
i = 1
if ends[starts.index(i)] == 0:
ends[starts.index(i)] = 1
pos.append([i - 1, ends[starts.index(i)]])
positions.append(pos)
starts = []
ends = []
elif "mRNA" not in line:
starts.append(int(columns[3]))
ends.append(int(columns[4]))
x += 1
if len(orfnames) == 0:
logging.error("GlimmerHMM gene prediction failed. Please check the " \
"format of your input FASTA file.")
#Create seq_record features for identified genes
idx = 0
for orfname in orfnames:
bpy_strand = strands[idx]
genepositions = positions[idx]
#For genes with only one CDS
if len(genepositions) == 1:
gstart, gend = genepositions[0]
loc = FeatureLocation(gstart, gend, strand=bpy_strand)
feature = SeqFeature(
location=loc,
id=orfname,
type="CDS",
qualifiers={
'locus_tag':
['ctg%s_%s' % (options.record_idx, orfname)]
})
seq_record.features.append(feature)
#For genes with multiple exons
else:
gstart, gend = min(genepositions[0]), max(genepositions[-1])
sublocations = []
for exonstart, exonend in genepositions:
exonloc = FeatureLocation(exonstart,
exonend,
strand=bpy_strand)
sublocations.append(exonloc)
loc = CompoundLocation(sublocations)
feature = SeqFeature(
location=loc,
id=orfname,
type="CDS",
qualifiers={
'locus_tag':
['ctg%s_%s' % (options.record_idx, orfname)]
})
seq_record.features.append(feature)
idx += 1
def write_to_genbank(self, filename: str = None, directory: str = None, record: SeqRecord = None) -> None:
""" Writes a genbank file containing only the information contained
within the Region.
"""
if not filename:
filename = "%s.region%03d.gbk" % (self.parent_record.id, self.get_region_number())
if directory:
filename = os.path.join(directory, filename)
if record is None:
record = self.parent_record.to_biopython()
assert isinstance(record, SeqRecord)
with warnings.catch_warnings():
warnings.simplefilter("ignore")
cluster_record = record[self.location.start:self.location.end]
cluster_record.annotations["date"] = record.annotations.get("date", '')
cluster_record.annotations["source"] = record.annotations.get("source", '')
cluster_record.annotations["organism"] = record.annotations.get("organism", '')
cluster_record.annotations["taxonomy"] = record.annotations.get("taxonomy", [])
cluster_record.annotations["data_file_division"] = record.annotations.get("data_file_division", 'UNK')
cluster_record.annotations["comment"] = record.annotations.get("comment", '')
# biopython does not persist the molecule_type annotation in slices,
# despite it being required for output to the genbank format
cluster_record.annotations["molecule_type"] = record.annotations["molecule_type"]
# update the antiSMASH annotation to include some cluster details
comment_end_marker = "##antiSMASH-Data-END"
cluster_comment = ("NOTE: This is a single cluster extracted from a larger record!\n"
"Orig. start :: {start}\n"
"Orig. end :: {end}\n"
"{end_marker}").format(start=self.location.start,
end=self.location.end,
end_marker=comment_end_marker)
original = cluster_record.annotations["comment"]
cluster_record.annotations["comment"] = original.replace(comment_end_marker, cluster_comment)
# our cut-out clusters are always linear
cluster_record.annotations["topology"] = "linear"
# renumber clusters, candidate_clusters and regions to reflect changes
# also update positions of RiPP component locations
if self.candidate_clusters:
first_candidate_cluster = min(sc.get_candidate_cluster_number() for sc in self.candidate_clusters)
first_cluster = min(cluster.get_protocluster_number() for cluster in self.get_unique_protoclusters())
else:
first_candidate_cluster = 0
first_cluster = 0
first_subregion = min(sub.get_subregion_number() for sub in self.subregions) if self.subregions else 0
for feature in cluster_record.features:
if feature.type == Region.FEATURE_TYPE:
candidates = feature.qualifiers.get("candidate_cluster_numbers")
if not candidates:
continue
candidates = [str(int(num) - first_candidate_cluster + 1) for num in candidates]
feature.qualifiers["candidate_cluster_numbers"] = candidates
elif feature.type == CandidateCluster.FEATURE_TYPE:
new = str(int(feature.qualifiers["candidate_cluster_number"][0]) - first_candidate_cluster + 1)
feature.qualifiers["candidate_cluster_number"] = [new]
new_clusters = [str(int(num) - first_cluster + 1) for num in feature.qualifiers["protoclusters"]]
feature.qualifiers["protoclusters"] = new_clusters
elif feature.type in ["protocluster", "proto_core"]:
new = str(int(feature.qualifiers["protocluster_number"][0]) - first_cluster + 1)
feature.qualifiers["protocluster_number"] = [new]
elif feature.type == "subregion":
new = str(int(feature.qualifiers["subregion_number"][0]) - first_subregion + 1)
feature.qualifiers["subregion_number"] = [new]
elif feature.type == "CDS_motif":
for qual in ["leader_location", "tail_location"]:
if qual not in feature.qualifiers:
continue
loc = location_from_string(feature.qualifiers[qual][0])
parts = []
for part in loc.parts:
new_start = part.start - self.location.start
new_end = part.end - self.location.start
parts.append(FeatureLocation(new_start, new_end, part.strand))
feature.qualifiers[qual] = [str(build_location_from_others(parts))]
seqio.write([cluster_record], filename, 'genbank')
