def CmalignScoreParser(lines):
"""Parser for tabfile format cmalign score result.
- IMPORTANT: Will only parse standard output from cmalign.
- NOTE: Will only work with search result files with a single CM
as a query. Will not work with multiple alignment result files
that have been concatenated.
- Result will be list of hits with following order:
[seq idx, seq name, seq len, total bit score, struct bit score,
avg prob, elapsed time]
"""
# Converting indices and %GC to integers and bit score to float.
# Since E-value is only present if CM is calibrated, leaving as string.
conversion_fields = [(0,int),(2,int),(3,float),(4,float),(5,float)]
cmalign_score_converter = ConvertFields(conversion_fields)
#Ignore hash characters
good_lines = []
for l in lines:
line = l.strip()
if line.startswith('# STOCKHOLM 1.0'):
break
if line and (not line.startswith('#')):
good_lines.append(l)
#make parser
cmalign_score_parser = SeparatorFormatParser(with_header=False,\
converter=cmalign_score_converter,\
ignore=None,\
sep=None)
return cmalign_score_parser(good_lines)
def CmsearchParser(lines):
"""Parser for tabfile format cmsearch result.
- IMPORTANT: Will not parse standard output from cmsearch. You must
use --tabfile with cmsearch to get correct format to use this
parser.
- NOTE: Will only work with search result files with a single CM
as a query. Will not work with multiple search result files
that have been concatenated.
- Result will be list of hits with following order:
[target name, target start, target stop, query start, query stop,
bit score, E-value, GC%]
"""
# Converting indices and %GC to integers and bit score to float.
# Since E-value is only present if CM is calibrated, leaving as string.
conversion_fields = [(2,int),(3,int),(4,int),(5,int),(6,float),(8,int)]
cmsearch_converter = ConvertFields(conversion_fields)
#Ignore hash characters
good_lines = []
for l in lines:
if not l.startswith('#'):
good_lines.append(l)
#make parser
cmsearch_parser = SeparatorFormatParser(with_header=False,\
converter=cmsearch_converter,\
ignore=None,\
sep=None)
return cmsearch_parser(good_lines)
line.append('')
header = []
for t, b in zip(*lines):
if t.strip().endswith('-'):
c = t.strip() + b
else:
c = ' '.join([t.strip(), b.strip()])
header += [c.strip()]
return header
int_series = lambda x: map(int, x.replace(',', ' ').split())
row_converter = ConvertFields([(i, int) for i in range(8)]+\
[(i, int) for i in range(10, 13)]+\
[(i, int) for i in range(14, 18)]+\
[(i, int_series) for i in range(18, 21)])
def MinimalPslParser(data, row_converter=row_converter):
"""returns version, header and rows from data"""
if type(data) == str:
data = open(data)
psl_version = None
header = None
rows = []
for record in data:
if psl_version is None:
assert 'psLayout version' in record
"""
from cogent import LoadTable
from cogent.parse.table import ConvertFields
__author__ = "Gavin Huttley, Anuj Pahwa"
__copyright__ = "Copyright 2007-2016, The Cogent Project"
__credits__ = ["Rob Knight", "Peter Maxwell", "Gavin Huttley", "Anuj Pahwa"]
__license__ = "GPL"
__version__ = "1.9"
__maintainer__ = "Gavin Huttley"
__email__ = "*****@*****.**"
__status__ = "Development"
# The 4th and the 7th elements of the row of data returned from bowtie are
# integer values and can thus be converted.
row_converter = ConvertFields([(3, int), (6, int)])
def BowtieOutputParser(data, row_converter=row_converter):
"""yields a header and row of data from the default bowtie output
Arguments:
- row_converter: if not provided, uses a default converter which casts
the Offset and Other Matches fields to ints. If set to None, all
returned data will be strings (this is faster).
"""
header = [
'Query Name', 'Strand Direction', 'Reference Name', 'Offset',
'Query Seq', 'Quality', 'Other Matches', 'Mismatches'
]
return val
def get_strand(val):
"""returns 1/-1 for strand from bitwise operation"""
v = int(val)
strand = [-1, 1][v & 16 == 0]
return strand
def zero_based(val):
"""returns a zero-based integer"""
return int(val) - 1
strict_converter = ConvertFields([(1, int), (3, int), (4, int),
(5, _strict_cigar_span)])
converter = ConvertFields([(1, get_strand), (3, zero_based), (4, _int_str),
(5, _cigar_span)])
# SAM fields: QNAME, FLAG, RNAME, POS, MAPQ, CIGAR, RNEXT, PNEXT, TLEN, SEQ, QUAL, OPT
complete_converter = ConvertFields([(0, str), (1, get_strand), (2, str),
(3, zero_based), (4, int),
(5, _cigar_span), (6, str), (7, int),
(8, int), (9, str), (10, str), (11, str)])
def MinimalSamParser(data, converter=converter):
"""returns records from a sam file
NOTE: the default converter turns the 1-based numbering of POS into
""" returns 1/-1 for strand in place of '+' or '-' """
strand = [-1, 1][val == '+']
return strand
pattern = re.compile(r'[0-9,X,Y,MT]+')
def _get_chrom(val):
""" returns the int component of a chromosome number """
chrom = pattern.search(val).group(0)
return chrom
# BED3 defines: chrom, chromStart, chromEnd
bed3_converter = ConvertFields([(0, _get_chrom), (1, int), (2, int)])
# BED6 adds: Name, score, strand
converter = ConvertFields([(0, _get_chrom), (1, int), (2, int), (3, str),
(4, int), (5, _get_strand)])
# BED12 additional fields: thickStart, thickEnd, itemRgb, blockCount, blockSizes, blockStarts
complete_converter = ConvertFields([(0, _get_chrom), (1, int), (2, int),
(3, str), (4, int), (5, _get_strand),
(6, int), (7, int), (8, tuple), (9, int),
(10, tuple), (11, tuple)])
def MinimalBedParser(data, converter=converter):
"""returns data lines from a BED file