def hits(self):
from iotbx import bioinformatics
for h in self.root.hits:
pieces = h.identifier.split("_")
assert 0 < len(pieces)
if len(pieces) == 1:
pdb = pieces[0]
chain = ""
else:
(pdb, chain) = pieces[:2]
alignment = bioinformatics.clustal_alignment(
names=["target", "%s_%s" % (pdb, chain)],
alignments=[
h.alignments[0].query.seq, h.alignments[0].match.seq
],
program="WU-BLAST")
yield bioinformatics.homology_search_hit(identifier=pdb,
chain=chain,
annotation=h.description,
alignment=alignment)
def hits(self):
names = self.alignment.names[ 1 : self.max_count ]
alignments = self.alignment.alignments[ 1 : self.max_count ]
try:
descriptions = self.alignment.descriptions[ 1 : self.max_count ]
except AttributeError:
descriptions = alignments
base = self.alignment.alignments[0]
for ( n, seq, desc ) in zip( names, alignments, descriptions ):
pieces = n.split( "_" )
assert 0 < len( pieces )
if len( pieces ) == 1:
pdb = pieces[0]
chain = ""
else:
( pdb, chain ) = pieces[:2]
alignment = bioinformatics.clustal_alignment(
names = [ "target", "%s_%s" % ( pdb, chain ) ],
alignments = [ base, seq ],
program = "<unknown>"
)
yield bioinformatics.homology_search_hit(
identifier = pdb,
chain = chain,
annotation = desc,
alignment = alignment
)
def get_muscle_alignment_ordered(sequences, out = None):
from iotbx import bioinformatics
name_for = {}
for ( i, seq ) in enumerate( sequences, start = 1 ):
name = name_for.get( seq, "Chain_%d" % i )
name_for[ seq ] = name
alignment = get_muscle_alignment(
fasta_sequences = "\n".join(
str( bioinformatics.sequence( name = name, sequence = seq.sequence ) )
for ( seq, name ) in name_for.items()
),
out = out,
)
lookup = dict( zip( alignment.names, alignment.alignments ) )
assert all( n in lookup for n in name_for.values() )
return bioinformatics.clustal_alignment(
names = [ seq.name for seq in sequences ],
alignments = [ lookup[ name_for[ seq ] ] for seq in sequences ],
program = alignment.program
)
def hits(self):
if not self.root.iterations:
raise StopIteration
from iotbx import bioinformatics
for h in self.root.iterations[-1].hits:
pieces = h.accession.split("_")
assert 0 < len(pieces)
if len(pieces) == 1:
pdb = pieces[0]
chain = ""
else:
(pdb, chain) = pieces[:2]
alignment = bioinformatics.clustal_alignment(
names=["target", "%s_%s" % (pdb, chain)],
alignments=[h.hsps[0].query.seq, h.hsps[0].hit.seq],
program="NCBI-BLAST")
yield bioinformatics.homology_search_hit(identifier=pdb,
chain=chain,
annotation=h.annotation,
alignment=alignment)
def hits(self):
if not self.root.iterations:
raise StopIteration
from iotbx import bioinformatics
for h in self.root.iterations[-1].hits:
pieces = h.accession.split( "_" )
assert 0 < len( pieces )
if len( pieces ) == 1:
pdb = pieces[0]
chain = ""
else:
( pdb, chain ) = pieces[:2]
alignment = bioinformatics.clustal_alignment(
names = [ "target", "%s_%s" % ( pdb, chain ) ],
alignments = [ h.hsps[0].query.seq, h.hsps[0].hit.seq ],
program = "NCBI-BLAST"
)
yield bioinformatics.homology_search_hit(
identifier = pdb,
chain = chain,
annotation = h.annotation,
alignment = alignment
)
def hits(self):
from iotbx import bioinformatics
for h in self.root.hits:
pieces = h.identifier.split( "_" )
assert 0 < len( pieces )
if len( pieces ) == 1:
pdb = pieces[0]
chain = ""
else:
( pdb, chain ) = pieces[:2]
alignment = bioinformatics.clustal_alignment(
names = [ "target", "%s_%s" % ( pdb, chain ) ],
alignments = [ h.alignments[0].query.seq, h.alignments[0].match.seq ],
program = "WU-BLAST"
)
yield bioinformatics.homology_search_hit(
identifier = pdb,
chain = chain,
annotation = h.description,
alignment = alignment
)
def get_muscle_alignment_ordered(sequences, out = None):
from iotbx import bioinformatics
from iotbx.pdb.amino_acid_codes import validate_sequence
name_for = {}
for ( i, seq ) in enumerate( sequences, start = 1 ):
name = name_for.get( seq, "Chain_%d" % i )
name_for[ seq ] = name
alignment, errors = get_muscle_alignment(
fasta_sequences = "\n".join(
str( bioinformatics.sequence( name = name, sequence = seq.sequence ) )
for ( seq, name ) in name_for.items()
),
out = out,
)
# check for errors and handle:
# invalid characters in sequences
if (len(errors) > 0):
for error in errors:
error = error.strip()
if ('Invalid character' in error):
for seq in name_for.keys():
invalid = validate_sequence(
seq.sequence, protein=True, strict_protein=False,
nucleic_acid=True, strict_nucleic_acid=False)
if (len(invalid) > 0):
name_for.pop(seq)
sequences = name_for.keys()
elif (len(error) > 0):
raise Sorry(error)
lookup = dict( zip( alignment.names, alignment.alignments ) )
assert all( n in lookup for n in name_for.values() )
return bioinformatics.clustal_alignment(
names = [ seq.name for seq in sequences ],
alignments = [ lookup[ name_for[ seq ] ] for seq in sequences ],
program = alignment.program
)