def rmsd(target, source, fragments, pdb, save=None):
matches = []
try:
src_file = wwpdb.find(source, [pdb])
if src_file is None:
raise IOError("Can't find structure {0} in {1}".format(
source, pdb))
src_structure = wwpdb.RegularStructureParser(
src_file).parse_structure()
except (IOError, structure.Broken3DStructureError,
wwpdb.PDBParseError) as ex:
matches.append('Error parsing {0:5}: {1!s}'.format(source, ex))
return matches
for fn, fragment in enumerate(fragments):
try:
if fragment.chain not in ('_', '', None):
src_chain = src_structure.chains[fragment.chain]
else:
src_chain = src_structure.first_chain
try:
query = target.subregion(fragment.qstart, fragment.qend)
subject = src_chain.subregion(fragment.start,
fragment.end,
clone=True)
except IndexError:
matches.append(
'Fragment {1.source_id:>5} {0.start:>4}-{0.end:>4} is out of range'
.format(fragment))
continue
si = query.align(subject)
match = csb.bio.fragments.FragmentMatch(fragment.id,
qstart=fragment.qstart,
qend=fragment.qend,
probability=None,
rmsd=si.rmsd,
tm_score=None,
qlength=target.length)
matches.append(match)
if save:
dummy = structure.Structure(subject.entry_id)
dummy.chains.append(subject)
filename = '{0.qstart:}-{0.qend}-{0.source_id}.{1.entry_id}{2}.frag'.format(
fragment, query, fn or '')
dummy.to_pdb(os.path.join(save, filename))
except (structure.Broken3DStructureError, IOError) as ex:
matches.append("Can't superimpose fragment {0}: {1!s}".format(
fragment.id, ex))
continue
return matches
def testAppend(self):
testModel = structure.Structure('test')
testModel.model_id = 3
rank = self.ensemble.models.append(testModel)
self.assertEqual(rank, 3)
self.assertEqual(self.ensemble.models[3], testModel)
self.assertEqual(self.ensemble.models[3].model_id, 3)
self.assertRaises(Exception, self.ensemble.models.append, 1)
def write_pdb(self, coords, filename):
if coords.ndim == 2: coords = coords.reshape(1, -1, 3)
ensemble = structure.Ensemble()
for i, xyz in enumerate(coords, 1):
chain = make_chain(xyz)
struct = structure.Structure('')
struct.chains.append(chain)
struct.model_id = i
ensemble.models.append(struct)
ensemble.to_pdb(filename)
def format_structure(self, input_pdb, chain_id, output_file):
"""
Format input_pdb as an HHpred pre-parsed structure file for the current
HMM. Formatting means: only chain_id left in the PDB file, residue
sequence numbers strictly corresponding to the HMM states' ranks.
@param input_pdb: source PDB file name
@type input_pdb: str
@param chain_id: source chain ID (which chain to pick from the
structure)
@type chain_id: str
@param output_file: save the output PDB file here
@type output_file: str
@raise StructureFormatError: when the specified PDB chain does not
correspond to the HMM.
"""
hmm = self.parse()
s = csb.bio.io.StructureParser(input_pdb).parse_structure()
chain = s.chains[chain_id]
if chain.length != hmm.layers.length:
raise StructureFormatError(
"{0}: Incorrect number of residues".format(chain.entry_id))
for layer, residue in zip(hmm.layers, chain.residues):
if residue.type == ProteinAlphabet.UNK:
residue.type = layer.residue.type
if residue.type != layer.residue.type:
msg = "Expected residue of type {0} at position {1}"
raise StructureFormatError(
msg.format(layer.residue.type, layer.rank))
residue.id = str(residue.rank), '$'
for residue in chain.residues:
residue.id = str(residue.rank), None
new_structure = structure.Structure(hmm.id)
new_structure.chains.append(chain.clone())
new_structure.to_pdb(output_file)