def main(gff_file=None,
fasta_file=None,
embedded_fasta=False,
stype=None,
user_defined=None,
dline=None,
qc=True,
output_prefix=None,
logger=None):
stderr_handler = logging.StreamHandler()
stderr_handler.setFormatter(
logging.Formatter('%(levelname)-8s %(message)s'))
logger_null = logging.getLogger(__name__ + 'null')
null_handler = logging.NullHandler()
logger_null.addHandler(null_handler)
if not gff_file or (not fasta_file and not embedded_fasta) or not stype:
print(
'Gff file, fasta file, and type of extracted sequences need to be specified'
)
sys.exit(1)
type_set = [
'gene', 'exon', 'pre_trans', 'trans', 'cds', 'pep', 'all',
'user_defined'
]
if not stype in type_set:
logger.error(
'Your sequence type is "{0:s}". Sequence type must be one of {1:s}!'
.format(stype, str(type_set)))
sys.exit(1)
if stype == 'all' and output_prefix:
pass
elif stype != 'all' and output_prefix:
logger.info('Specifying prefix of output file name: (%s)...',
output_prefix)
fname = '{0:s}_{1:s}.fa'.format(output_prefix, stype)
report_fh = open(fname, 'w')
else:
print('[Error] Please specify the prefix of output file name...')
sys.exit(1)
if stype == 'user_defined' and user_defined != None:
if len(user_defined) != 2:
logger.error(
'Please specify parent and child feature via the -u argument. Format: [parent feature type],[child feature type]'
)
sys.exit(1)
elif stype != 'user_defined' and user_defined != None:
logger.warning(
'Your sequence type is "{0:s}", -u argument will be ignored.'.
format(stype))
elif stype == 'user_defined' and user_defined == None:
logger.error('-u is needed in combination with -st user_defined.')
sys.exit(1)
logger.info('Reading files: {0:s}, {1:s}...'.format(gff_file, fasta_file))
gff = None
if qc:
initial_phase = False
gff = Gff3(gff_file=gff_file, fasta_external=fasta_file, logger=logger)
if embedded_fasta and len(gff.fasta_embedded) == 0:
logger.error('There is no embedded fasta in the GFF3 file.')
sys.exit(1)
logger.info('Checking errors...')
gff.check_parent_boundary()
gff.check_phase(initial_phase)
gff.check_reference()
error_set = function4gff.extract_internal_detected_errors(gff)
t = intra_model.main(gff, logger=logger)
if t:
error_set.extend(t)
t = single_feature.main(gff, logger=logger)
if t:
error_set.extend(t)
if error_set and len(error_set):
escaped_error = ['Esf0012', 'Esf0033']
eSet = list()
for e in error_set:
if not e['eCode'] in escaped_error:
eSet.append(e)
if len(eSet):
logger.warning(
'The extracted sequences might be wrong for the following features which have formatting errors...'
)
print('ID\tError_Code\tError_Tag')
for e in eSet:
tag = '[{0:s}]'.format(e['eTag'])
print(e['ID'], e['eCode'], tag)
else:
gff = Gff3(gff_file=gff_file,
fasta_external=fasta_file,
logger=logger_null)
if embedded_fasta and len(gff.fasta_embedded) == 0:
logger.error('There is no embedded fasta in the GFF3 file.')
logger.info('Extract sequences for {0:s}...'.format(stype))
seq = dict()
if stype == 'all':
if output_prefix:
logger.info('Specifying prefix of output file name: (%s)...',
output_prefix)
pass
else:
print('[Error] Please specify the prefix of output file name...')
sys.exit(1)
tmp_stype = 'pre_trans'
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
seq = extract_start_end(gff, tmp_stype, dline, embedded_fasta)
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
seq = dict()
tmp_stype = 'gene'
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
seq = extract_start_end(gff, tmp_stype, dline, embedded_fasta)
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
seq = dict()
tmp_stype = 'exon'
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
seq = extract_start_end(gff, tmp_stype, dline, embedded_fasta)
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
seq = dict()
tmp_stype = 'trans'
feature_type = ['exon', 'pseudogenic_exon']
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
seq = splicer(gff, feature_type, dline, stype, embedded_fasta)
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
seq = dict()
tmp_stype = 'cds'
feature_type = ['CDS']
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
seq = splicer(gff, feature_type, dline, stype, embedded_fasta)
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
seq = dict()
tmp_stype = 'pep'
feature_type = ['CDS']
logger.info('\t- Extract sequences for {0:s}...'.format(tmp_stype))
tmpseq = splicer(gff, feature_type, dline, tmp_stype, embedded_fasta)
for k, v in tmpseq.items():
k = k.replace("|mRNA(CDS)|", "|peptide|")
v = translator(v)
seq[k] = v
if len(seq):
fname = '{0:s}_{1:s}.fa'.format(output_prefix, tmp_stype)
report_fh = open(fname, 'w')
logger.info(
'\t\tPrint out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, tmp_stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
elif stype == 'user_defined':
feature_type = [user_defined[0], user_defined[1]]
seq = splicer(gff, feature_type, dline, stype, embedded_fasta)
if len(seq):
logger.info(
'Print out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
else:
if stype == 'pre_trans' or stype == 'gene' or stype == 'exon':
seq = extract_start_end(gff, stype, dline, embedded_fasta)
elif stype == 'trans':
feature_type = ['exon', 'pseudogenic_exon']
seq = splicer(gff, feature_type, dline, stype, embedded_fasta)
elif stype == 'cds':
feature_type = ['CDS']
seq = splicer(gff, feature_type, dline, stype, embedded_fasta)
elif stype == 'pep':
feature_type = ['CDS']
tmpseq = splicer(gff, feature_type, dline, stype, embedded_fasta)
for k, v in tmpseq.items():
k = k.replace("|mRNA(CDS)|", "|peptide|")
#k = re.sub(r'(.*-)(R)(.)',r'\1P\3',k)
v = translator(v)
seq[k] = v
if len(seq):
logger.info(
'Print out extracted sequences: {0:s}_{1:s}.fa...'.format(
output_prefix, stype))
for k, v in seq.items():
if len(k) != 0 and len(v) != 0:
report_fh.write('{0:s}\n{1:s}\n'.format(k, v))
def script_main():
logger_stderr = logging.getLogger(__name__ + 'stderr')
logger_stderr.setLevel(logging.INFO)
stderr_handler = logging.StreamHandler()
stderr_handler.setFormatter(
logging.Formatter('%(levelname)-8s %(message)s'))
logger_stderr.addHandler(stderr_handler)
logger_null = logging.getLogger(__name__ + 'null')
null_handler = logging.NullHandler()
logger_null.addHandler(null_handler)
import argparse
from textwrap import dedent
parser = argparse.ArgumentParser(
formatter_class=argparse.RawDescriptionHelpFormatter,
description=dedent("""\
Testing environment:
1. Python 2.7
Inputs:
1. GFF3: Specify the file name with the -g or --gff argument; Please note that this program requires gene/pseudogene and mRNA/pseudogenic_transcript to have an ID attribute in column 9.
2. fasta file: Specify the file name with the -f or --fasta argument
Outputs:
1. Error report for the input GFF3 file
* Line_num: Line numbers of the found problematic models in the input GFF3 file.
* Error_code: Error codes for the found problematic models. Please refer to lib/ERROR/ERROR.py to see the full list of Error_code and the corresponding Error_tag.
* Error_tag: Detail of the found errors for the problematic models. Please refer to lib/ERROR/ERROR.py to see the full list of Error_code and the corresponding Error_tag.
Quick start:
gff3_QC -g example_file/example.gff3 -f example_file/reference.fa -o test
or
gff3_QC --gff example_file/example.gff3 --fasta example_file/reference.fa --output test
"""))
parser.add_argument('-g',
'--gff',
type=str,
help='Genome annotation file, gff3 format')
parser.add_argument('-f',
'--fasta',
type=str,
help='Genome sequences, fasta format')
parser.add_argument(
'-noncg',
'--noncanonical_gene',
action="store_true",
help='gff3 file is not formatted in the canonical gene model format.')
parser.add_argument(
'-i',
'--initial_phase',
action="store_true",
help='Check whether initial CDS phase is 0 (default: no check)')
parser.add_argument(
'-n',
'--allowed_num_of_n',
type=int,
default=0,
help=
'Max number of Ns allowed in a feature, anything more will be reported as an error (default: 0)'
)
parser.add_argument(
'-t',
'--check_n_feature_types',
nargs='*',
default=['CDS'],
help=
'Count the number of Ns in each feature with the type specified, multiple types may be specified, ex: -t CDS exon (default: "CDS")'
)
parser.add_argument('-o',
'--output',
type=str,
help='output file name (default: report.txt)')
parser.add_argument('-s',
'--statistic',
type=str,
help='statistic file name (default: statistic.txt)')
parser.add_argument('-v',
'--version',
action='version',
version='%(prog)s ' + __version__)
args = parser.parse_args()
if args.gff:
logger_stderr.info('Checking gff file (%s)...', args.gff)
elif not sys.stdin.isatty(): # if STDIN connected to pipe or file
args.gff = sys.stdin
logger_stderr.info('Reading from STDIN...')
else: # no input
parser.print_help()
sys.exit(1)
if args.fasta:
logger_stderr.info('Checking genome fasta (%s)...', args.fasta)
elif not sys.stdin.isatty(): # if STDIN connected to pipe or file
args.fasta = sys.stdin
logger_stderr.info('Reading from STDIN...')
else: # no input
parser.print_help()
sys.exit(1)
if args.allowed_num_of_n or args.check_n_feature_types:
check_n = True
else:
check_n = False
logger_stderr.info('Reading gff files: (%s)...\n', args.gff)
gff3 = Gff3(gff_file=args.gff,
fasta_external=args.fasta,
logger=logger_null)
logger_stderr.info('Checking errors in the gff files: (%s)...\n', args.gff)
if not gff3.check_parent_boundary():
sys.exit()
gff3.check_unresolved_parents()
if args.noncanonical_gene == False:
gff3.check_phase(args.initial_phase)
gff3.check_reference(fasta_external=args.fasta,
check_n=check_n,
allowed_num_of_n=args.allowed_num_of_n,
feature_types=args.check_n_feature_types)
logger_stderr.info('\t- Checking missing attributes: (%s)...\n',
'function4gff.FIX_MISSING_ATTR()')
function4gff.FIX_MISSING_ATTR(gff3, logger=logger_stderr)
error_set = list()
cmd = None
cmd = function4gff.extract_internal_detected_errors(gff3)
if cmd:
error_set.extend(cmd)
cmd = None
logger_stderr.info('\t- Checking intra-model errors: (%s)...\n', args.gff)
cmd = intra_model.main(gff3,
logger=logger_stderr,
noncanonical_gene=args.noncanonical_gene)
if cmd:
error_set.extend(cmd)
cmd = None
logger_stderr.info('\t- Checking inter-model errors: (%s)...\n', args.gff)
cmd = inter_model.main(gff3,
args.gff,
args.fasta,
logger=logger_stderr,
noncanonical_gene=args.noncanonical_gene)
if cmd:
error_set.extend(cmd)
cmd = None
logger_stderr.info('\t- Checking single-feature errors: (%s)...\n',
args.gff)
cmd = single_feature.main(gff3, logger=logger_stderr)
if cmd:
error_set.extend(cmd)
if args.output:
logger_stderr.info('Print QC report at {0:s}'.format(args.output))
report_fh = open(args.output, 'w')
else:
logger_stderr.info('Print QC report at {0:s}'.format('report.txt'))
report_fh = open('report.txt', 'w')
if args.statistic:
logger_stderr.info('Print QC statistic report at {0:s}'.format(
args.statistic))
statistic_fh = open(args.statistic, 'w')
else:
logger_stderr.info(
'Print QC statistic report at {0:s}'.format('statistic.txt'))
statistic_fh = open('statistic.txt', 'w')
report_fh.write('Line_num\tError_code\tError_tag\n')
for e in sorted(error_set, key=lambda x: sorted(x.keys())):
tag = '[{0:s}]'.format(e['eTag'])
report_fh.write('{0:s}\t{1:s}\t{2:s}\n'.format(str(e['line_num']),
str(e['eCode']),
str(tag)))
#statistic_file
error_counts = dict()
ERROR_INFO = ERROR.INFO
statistic_fh.write('Error_code\tNumber_of_problematic_models\tError_tag\n')
for s in sorted(error_set, key=lambda x: sorted(x.keys())):
if s['eCode'] not in error_counts:
error_counts[s['eCode']] = {
'count': 0,
'etag': ERROR_INFO[s['eCode']]
}
error_counts[s['eCode']]['count'] += 1
for a in error_counts:
statistic_fh.write('{0:s}\t{1:s}\t{2:s}\n'.format(
str(a), str(error_counts[a]['count']),
str(error_counts[a]['etag'])))