def renumber_pdb_needleman(inseq, instr, seqback=False):
"""Return modified :class:`gemmi.Structure` with new residue numbers. It uses Needleman-Wunsch algorithm to perform the sequence alignment.
:param inseq: Input sequence.
:type inseq: :class:`crops.elements.sequences.oligoseq`
:param instr: Gemmi structure.
:type instr: :class:`gemmi.Structure`
:param seqback: If True, it additionally returns the :class:`crops.elements.sequences.oligoseq` with the gaps found in the structure, defaults to False.
:type seqback: bool, optional
:return instr: Renumbered structure.
:rtype instr: :class:`gemmi.Structure`
:return inseq: Sequence with extra information about gaps, only if seqback is True.
:rtype inseq: :class:`crops.elements.sequences.oligoseq`
"""
renumbered_structure = gemmi.Structure()
renumbered_structure.name = instr.name
if seqback:
for monkey in inseq.imer:
inseq.imer[monkey].seqs['gapseq'] = []
for model in instr:
renumbered_model = gemmi.Model(model.name)
for chain in model:
renumbered_chain = gemmi.Chain(chain.name)
nseq = inseq.whatseq(chain.name)
original_seq = inseq.imer[nseq].seqs['mainseq']
model_seq = ''.join([ressymbol(x.name) for x in chain])
aligned_dict = get_sequence_alignment(original_seq, model_seq)
if aligned_dict is None:
logging.warning('Alignment failed for chain {}, it will be excluded'.format(chain.name))
continue
for index, residue in enumerate(chain):
_residue = residue.clone()
_residue.seqid.num = aligned_dict[index] + 1
renumbered_chain.add_residue(_residue)
renumbered_model.add_chain(renumbered_chain)
if seqback:
res_set = {residue.seqid.num for residue in renumbered_chain}
newseq = ''
for n in range(len(original_seq)):
if (n+1) in res_set:
newseq += original_seq[n]
else:
newseq += '-'
inseq.imer[nseq].seqs['gapseq'].append(newseq)
renumbered_structure.add_model(renumbered_model)
if seqback:
return renumbered_structure, inseq
else:
return renumbered_structure
def test_model_all(self):
model = gemmi.Model('1')
for name in 'ABCDEFG':
model.add_chain(gemmi.Chain(name))
expected = []
for cname in 'BCF':
chain = model[cname]
for _ in range(7):
chain.add_residue(gemmi.Residue())
for (r, name) in [(2, '0'), (2, '1'), (3, '2'), (5, '3')]:
a = gemmi.Atom()
a.name = cname + name
expected.append(a.name)
chain[r].add_atom(a)
self.assertEqual([cra.atom.name for cra in model.all()], expected)
st = gemmi.Structure()
st.add_model(model)
st.remove_empty_chains()
self.assertEqual([cra.atom.name for cra in model.all()], expected)
def test_remove2(self):
model = gemmi.read_structure(full_path('1pfe.cif.gz'))[0]
self.assertEqual(len(model), 2)
del model['A']
self.assertEqual(len(model), 1)
b = model['B']
self.assertEqual(b[0].name, 'DSN')
del b['1']['DSN']
self.assertEqual(b[0].name, 'ALA')
del b[0]
self.assertEqual(b[0].name, 'N2C')
# test append_residues()
self.assertEqual(len(b), 20)
b.append_residues(b[:5], min_sep=10)
self.assertEqual(len(b), 25)
# test append_residues() with empty chain
new_chain = gemmi.Chain('X')
new_chain.append_residues(b[:5], min_sep=1)
self.assertEqual(len(new_chain), 5)
def get_masked_pdb(pdb: gemmi.Structure,
coord: Coord,
radius: float = 8.0) -> gemmi.Structure:
event_centoid = gemmi.Position(
coord.x,
coord.y,
coord.z,
)
new_structure = gemmi.Structure()
for model_i, model in enumerate(pdb):
new_model = gemmi.Model(model.name)
new_structure.add_model(new_model, pos=-1)
for chain_i, chain in enumerate(model):
new_chain = gemmi.Chain(chain.name)
new_structure[model_i].add_chain(new_chain, pos=-1)
for residue_i, residue in enumerate(chain):
new_residue = gemmi.Residue()
new_residue.name = residue.name
new_residue.seqid = residue.seqid
new_residue.subchain = residue.subchain
new_residue.label_seq = residue.label_seq
new_residue.het_flag = residue.het_flag
new_structure[model_i][chain_i].add_residue(new_residue,
pos=-1)
for atom_i, atom in enumerate(residue):
pos = atom.pos
if pos.dist(event_centoid) > radius:
new_structure[model_i][chain_i][residue_i].add_atom(
atom, pos=-1)
for model_i, model in enumerate(pdb):
pdb.add_model(new_structure[model_i], pos=-1)
del pdb[0]
return pdb
def test_remove2(self):
st = gemmi.read_structure(full_path('1pfe.cif.gz'))
model = st[0]
self.assertEqual(len(model), 2)
b = model['B']
self.assertEqual(b[0].name, 'DSN')
del b['1']['DSN']
self.assertEqual(b[0].name, 'ALA')
del b[0]
self.assertEqual(b[0].name, 'N2C')
# test append_residues()
self.assertEqual(len(b), 20)
b.append_residues(b[:5], min_sep=10)
self.assertEqual(len(b), 25)
# test append_residues() with empty chain
new_chain = gemmi.Chain('X')
new_chain.append_residues(b[:5], min_sep=1)
self.assertEqual(len(new_chain), 5)
# test adding and removing chains
model.add_chain(new_chain, unique_name=False)
model.add_chain(new_chain, unique_name=True)
model.add_chain(new_chain)
self.assertEqual([chain.name for chain in model], list('ABXCX'))
del model[2:]
model.add_chain(new_chain, unique_name=True)
self.assertEqual([chain.name for chain in model], list('ABX'))
del model[-1]
del model['A']
self.assertEqual(len(model), 1)
self.assertEqual(model[0].name, 'B')
doc = st.make_mmcif_document()
ref_seq = doc[0].get_mmcif_category('_struct_ref_seq')
self.assertEqual(ref_seq['pdbx_strand_id'], ['B'])