def read_blast_alignment(ofile, srcDB, destDB, al=None, pipeline=None,
translateSrc=False, translateDest=False):
"""Apply sequence of transforms to read input from 'ofile'.
srcDB: database for finding query sequences from the blast input;
destDB: database for finding subject sequences from the blast input;
al, if not None, must be a writeable alignment object in which to
store the alignment intervals;
translateSrc=True forces creation of a TranslationDB representing
the possible 6-frames of srcDB (for blastx, tblastx);
translateDest=True forces creation of a TranslationDB representing
the possible 6-frames of destDB (for tblastn, tblastx).
If pipeline is not None, it must be a list of filter functions each
taking a single argument and returning an iterator or iterable result
object.
"""
p = BlastHitParser()
d = dict(id='src_id',
start='src_start',
stop='src_end',
ori='src_ori',
idDest='dest_id',
startDest='dest_start',
stopDest='dest_end',
oriDest='dest_ori')
if translateSrc:
srcDB = translationDB.get_translation_db(srcDB)
if translateDest:
destDB = translationDB.get_translation_db(destDB)
cti = CoordsToIntervals(srcDB, destDB, d)
alignedIvals = cti(p.parse_file(ofile))
if pipeline is None:
result = save_interval_alignment(alignedIvals, al)
else: # apply all the filters in our pipeline one by one
result = alignedIvals
for f in pipeline:
result = f(result)
return result
def __getitem__(self, query):
"""generate slices for all translations of the query """
# generate NLMSA for this single sequence
al = self(query)
# get the translation database for the sequence
tdb = translationDB.get_translation_db(query.db)
# run through all of the frames & find alignments.
slices = []
for trans_seq in tdb[query.id].iter_frames():
try:
slice = al[trans_seq]
except KeyError:
continue
if not isinstance(slice, EmptySlice):
slices.append(slice)
return slices