def extend_protein_model(pdb_hierarchy,
mon_lib_srv,
add_hydrogens=None,
selection=None):
"""
Rebuild a sidechain by substituting an ideal amino acid and rotating the
sidechain to match the old conformation as closely as possible.
Limited functionality:
1) Amino-acids only, 2) side chain atoms only.
3) Not terminii aware
4) Not aware of v2.3 vs v3.2 atom names e.g. HB1,HB2 vs HB2,HB3
"""
from mmtbx.monomer_library import idealized_aa
from mmtbx.rotamer import rotamer_eval
from scitbx.array_family import flex
ideal_dict = idealized_aa.residue_dict()
pdb_atoms = pdb_hierarchy.atoms()
if (selection is None):
selection = flex.bool(pdb_atoms.size(), True)
partial_sidechains = []
for chain in pdb_hierarchy.only_model().chains():
for residue_group in chain.residue_groups():
for residue in residue_group.atom_groups():
i_seqs = residue.atoms().extract_i_seq()
residue_sel = selection.select(i_seqs)
if (not residue.resname.lower() in ideal_dict.keys()): continue
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=residue,
mon_lib_srv=mon_lib_srv,
ignore_hydrogens=False)
if (len(missing_atoms) > 0):
all_h = list(set([s.strip()[0] for s in missing_atoms
])) in [['H'], ['D'], ['T']]
if (add_hydrogens is False and all_h): continue
partial_sidechains.append(residue)
for residue in partial_sidechains:
residue_elements = [
e.strip() for e in residue.atoms().extract_element()
]
res_key = residue.resname.lower()
if (add_hydrogens is None):
if ("H" in residue_elements): res_key += "_h"
if (add_hydrogens is True): res_key += "_h"
target_atom_group = ideal_dict[res_key].only_model().only_chain().\
only_residue_group().only_atom_group()
new_residue = extend_residue(residue=residue,
target_atom_group=target_atom_group,
mon_lib_srv=mon_lib_srv)
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=new_residue,
mon_lib_srv=mon_lib_srv,
ignore_hydrogens=False)
#assert len(missing_atoms) == 0, missing_atoms
rg = residue.parent()
rg.remove_atom_group(residue)
rg.append_atom_group(new_residue.detached_copy())
pdb_hierarchy.atoms().reset_i_seq()
pdb_hierarchy.atoms().reset_serial()
return len(partial_sidechains)
def check_missing_atom(pdb_filename):
pdb_inp = iotbx.pdb.input(file_name=pdb_filename)
pdb_hierarchy = pdb_inp.construct_hierarchy()
ideal_dict = idealized_aa.residue_dict()
pdb_atoms = pdb_hierarchy.atoms()
selection = flex.bool(pdb_atoms.size(), True)
partial_sidechains = []
for chain in pdb_hierarchy.only_model().chains():
for residue_group in chain.residue_groups():
if (residue_group.atom_groups_size() != 1): continue
for residue in residue_group.atom_groups():
i_seqs = residue.atoms().extract_i_seq()
residue_sel = selection.select(i_seqs)
if (not residue.resname.lower() in ideal_dict.keys()): continue
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=residue,
mon_lib_srv=mon_lib_server,
ignore_hydrogens=True)
if (len(missing_atoms) > 0):
return True
return False
def filter_before_build (
pdb_hierarchy,
fmodel,
geometry_restraints_manager,
selection=None,
params=None,
verbose=True,
log=sys.stdout) :
"""
Pick residues suitable for building alternate conformations - by default,
this means no MolProbity/geometry outliers, good fit to map, no missing
atoms, and no pre-existing alternates, but with significant difference
density nearby.
"""
from mmtbx.validation import molprobity
from mmtbx.rotamer import rotamer_eval
import mmtbx.monomer_library.server
from mmtbx import building
from iotbx.pdb import common_residue_names_get_class
from scitbx.array_family import flex
if (selection is None) :
selection = flex.bool(fmodel.xray_structure.scatterers().size(), True)
pdb_atoms = pdb_hierarchy.atoms()
assert (pdb_atoms.size() == fmodel.xray_structure.scatterers().size())
pdb_atoms.reset_i_seq()
full_validation = molprobity.molprobity(
pdb_hierarchy=pdb_hierarchy,
fmodel=fmodel,
geometry_restraints_manager=geometry_restraints_manager,
outliers_only=False,
rotamer_library="8000")
if (verbose) :
full_validation.show(out=log)
multi_criterion = full_validation.as_multi_criterion_view()
if (params is None) :
params = libtbx.phil.parse(filter_params_str).extract()
mon_lib_srv = mmtbx.monomer_library.server.server()
two_fofc_map, fofc_map = building.get_difference_maps(fmodel=fmodel)
residues = []
filters = params.discard_outliers
make_sub_header("Identifying candidates for building", out=log)
# TODO parallelize
for chain in pdb_hierarchy.only_model().chains() :
if (not chain.is_protein()) :
continue
for residue_group in chain.residue_groups() :
atom_groups = residue_group.atom_groups()
id_str = residue_group.id_str()
i_seqs = residue_group.atoms().extract_i_seq()
residue_sel = selection.select(i_seqs)
if (not residue_sel.all_eq(True)) :
continue
if (len(atom_groups) > 1) :
print >> log, " %s is already multi-conformer" % id_str
continue
atom_group = atom_groups[0]
res_class = common_residue_names_get_class(atom_group.resname)
if (res_class != "common_amino_acid") :
print >> log, " %s: non-standard residue" % id_str
continue
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=atom_group,
mon_lib_srv=mon_lib_srv,
ignore_hydrogens=True)
if (len(missing_atoms) > 0) :
# residues modeled as pseudo-ALA are allowed by default; partially
# missing sidechains are more problematic
if ((building.is_stub_residue(atom_group)) and
(not params.ignore_stub_residues)) :
pass
else :
print >> log, " %s: missing or incomplete sidechain" % \
(id_str, len(missing_atoms))
continue
validation = multi_criterion.get_residue_group_data(residue_group)
is_outlier = is_validation_outlier(validation, params)
if (is_outlier) :
print >> log, " %s" % str(validation)
continue
if (params.use_difference_map) :
i_seqs_no_hd = building.get_non_hydrogen_atom_indices(residue_group)
map_stats = building.local_density_quality(
fofc_map=fofc_map,
two_fofc_map=two_fofc_map,
atom_selection=i_seqs_no_hd,
xray_structure=fmodel.xray_structure,
radius=params.sampling_radius)
if ((map_stats.number_of_atoms_below_fofc_map_level() == 0) and
(map_stats.fraction_of_nearby_grid_points_above_cutoff()==0)) :
if (verbose) :
print >> log, " no difference density for %s" % id_str
continue
residues.append(residue_group.only_atom_group())
if (len(residues) == 0) :
raise Sorry("No residues passed the filtering criteria.")
print >> log, ""
print >> log, "Alternate conformations will be tried for %d residue(s):" % \
len(residues)
building.show_chain_resseq_ranges(residues, out=log, prefix=" ")
print >> log, ""
return residues
def extend_protein_model(pdb_hierarchy,
selection=None,
hydrogens=Auto,
max_atoms_missing=None,
log=None,
modify_segids=True,
prefilter_callback=None,
idealized_residue_dict=None,
skip_non_protein_chains=True):
"""
Replace all sidechains with missing non-hydrogen atoms in a PDB hierarchy.
"""
from mmtbx.monomer_library import idealized_aa
from mmtbx.rotamer import rotamer_eval
import mmtbx.monomer_library.server
from iotbx.pdb import common_residue_names_get_class
from scitbx.array_family import flex
if (prefilter_callback is not None):
assert hasattr(prefilter_callback, "__call__")
else:
prefilter_callback = lambda r: True
ideal_dict = idealized_residue_dict
if (ideal_dict is None):
ideal_dict = idealized_aa.residue_dict()
if (log is None): log = null_out()
mon_lib_srv = mmtbx.monomer_library.server.server()
pdb_atoms = pdb_hierarchy.atoms()
if (selection is None):
selection = flex.bool(pdb_atoms.size(), True)
partial_sidechains = []
for chain in pdb_hierarchy.only_model().chains():
if (not chain.is_protein()) and (skip_non_protein_chains):
print >> log, " skipping non-protein chain '%s'" % chain.id
continue
for residue_group in chain.residue_groups():
atom_groups = residue_group.atom_groups()
if (len(atom_groups) > 1):
print >> log, " %s %s has multiple conformations, skipping" % \
(chain.id, residue_group.resid())
continue
residue = atom_groups[0]
i_seqs = residue.atoms().extract_i_seq()
residue_sel = selection.select(i_seqs)
if (not residue_sel.all_eq(True)):
continue
if (idealized_residue_dict is None):
res_class = common_residue_names_get_class(residue.resname)
if (res_class != "common_amino_acid"):
print >> log, " skipping non-standard residue %s" % residue.resname
continue
else:
key = residue.resname.lower()
if (hydrogens == True):
key = key + "_h"
if (not key in idealized_residue_dict.keys()):
pass
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=residue,
mon_lib_srv=mon_lib_srv,
ignore_hydrogens=True)
if (len(missing_atoms) > 0):
print >> log, " missing %d atoms in %s: %s" % (len(
missing_atoms), residue.id_str(), ",".join(missing_atoms))
if ((max_atoms_missing is None)
or (len(missing_atoms) < max_atoms_missing)):
if (prefilter_callback(residue)):
partial_sidechains.append(residue)
for residue in partial_sidechains:
new_residue = extend_residue(residue=residue,
ideal_dict=ideal_dict,
hydrogens=hydrogens,
mon_lib_srv=mon_lib_srv,
match_conformation=True)
if (modify_segids):
for atom in new_residue.atoms():
atom.segid = "XXXX"
rg = residue.parent()
rg.remove_atom_group(residue)
rg.append_atom_group(new_residue.detached_copy())
pdb_hierarchy.atoms().reset_i_seq()
pdb_hierarchy.atoms().reset_serial()
return len(partial_sidechains)
def extend_protein_model (pdb_hierarchy,
selection=None,
hydrogens=Auto,
max_atoms_missing=None,
log=None,
modify_segids=True,
prefilter_callback=None,
idealized_residue_dict=None,
skip_non_protein_chains=True) :
"""
Replace all sidechains with missing non-hydrogen atoms in a PDB hierarchy.
"""
from mmtbx.monomer_library import idealized_aa
from mmtbx.rotamer import rotamer_eval
import mmtbx.monomer_library.server
from iotbx.pdb import common_residue_names_get_class
from scitbx.array_family import flex
if (prefilter_callback is not None) :
assert hasattr(prefilter_callback, "__call__")
else :
prefilter_callback = lambda r: True
ideal_dict = idealized_residue_dict
if (ideal_dict is None) :
ideal_dict = idealized_aa.residue_dict()
if (log is None) : log = null_out()
mon_lib_srv = mmtbx.monomer_library.server.server()
pdb_atoms = pdb_hierarchy.atoms()
if (selection is None) :
selection = flex.bool(pdb_atoms.size(), True)
partial_sidechains = []
for chain in pdb_hierarchy.only_model().chains() :
if (not chain.is_protein()) and (skip_non_protein_chains) :
print >> log, " skipping non-protein chain '%s'" % chain.id
continue
for residue_group in chain.residue_groups() :
atom_groups = residue_group.atom_groups()
if (len(atom_groups) > 1) :
print >> log, " %s %s has multiple conformations, skipping" % \
(chain.id, residue_group.resid())
continue
residue = atom_groups[0]
i_seqs = residue.atoms().extract_i_seq()
residue_sel = selection.select(i_seqs)
if (not residue_sel.all_eq(True)) :
continue
if (idealized_residue_dict is None) :
res_class = common_residue_names_get_class(residue.resname)
if (res_class != "common_amino_acid") :
print >> log, " skipping non-standard residue %s" % residue.resname
continue
else :
key = residue.resname.lower()
if (hydrogens == True) :
key = key + "_h"
if (not key in idealized_residue_dict.keys()) :
pass
missing_atoms = rotamer_eval.eval_residue_completeness(
residue=residue,
mon_lib_srv=mon_lib_srv,
ignore_hydrogens=True)
if (len(missing_atoms) > 0) :
print >> log, " missing %d atoms in %s: %s" % (len(missing_atoms),
residue.id_str(), ",".join(missing_atoms))
if ((max_atoms_missing is None) or
(len(missing_atoms) < max_atoms_missing)) :
if (prefilter_callback(residue)) :
partial_sidechains.append(residue)
for residue in partial_sidechains :
new_residue = extend_residue(residue=residue,
ideal_dict=ideal_dict,
hydrogens=hydrogens,
mon_lib_srv=mon_lib_srv,
match_conformation=True)
if (modify_segids) :
for atom in new_residue.atoms() :
atom.segid = "XXXX"
rg = residue.parent()
rg.remove_atom_group(residue)
rg.append_atom_group(new_residue.detached_copy())
pdb_hierarchy.atoms().reset_i_seq()
pdb_hierarchy.atoms().reset_serial()
return len(partial_sidechains)