def molprobity_score(self):
"""
Compute overall measure of (protein!) structure quality based on an
empirical formula combining clashscore, rotamer outliers, and Ramachandran
favored statistics. The result is on approximately the same scale as
resolution.
"""
if (None in [self.ramalyze, self.rotalyze, self.clashes]):
return None
from mmtbx.validation import utils
return utils.molprobity_score(clashscore=self.clashes.get_clashscore(),
rota_out=self.rotalyze.out_percent,
rama_fav=self.ramalyze.fav_percent)
def molprobity_score (self) :
"""
Compute overall measure of (protein!) structure quality based on an
empirical formula combining clashscore, rotamer outliers, and Ramachandran
favored statistics. The result is on approximately the same scale as
resolution.
"""
if (None in [self.ramalyze, self.rotalyze, self.clashes]) :
return None
from mmtbx.validation import utils
return utils.molprobity_score(
clashscore=self.clashes.get_clashscore(),
rota_out=self.rotalyze.out_percent,
rama_fav=self.ramalyze.fav_percent)
def __init__(
self,
pdb_hierarchy,
molprobity_scores=False,
):
self.pdb_hierarchy = pdb_hierarchy
self.clashscore = None
self.ramachandran_outliers = None
self.ramachandran_allowed = None
self.ramachandran_favored = None
self.rotamer_outliers = None
self.c_beta_dev = None
self.mpscore = None
self.omglz = None
self.n_cis_proline = None
self.n_cis_general = None
self.n_twisted_proline = None
self.n_twisted_general = None
if(molprobity_scores):
self.ramalyze_obj = ramalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.ramachandran_outliers = self.ramalyze_obj.percent_outliers
self.ramachandran_allowed = self.ramalyze_obj.percent_allowed
self.ramachandran_favored = self.ramalyze_obj.percent_favored
self.rotalyze_obj = rotalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.rotamer_outliers = self.rotalyze_obj.percent_outliers
self.cbetadev_obj = cbetadev(
pdb_hierarchy = pdb_hierarchy,
outliers_only = True,
out = null_out())
self.c_beta_dev = self.cbetadev_obj.get_outlier_count()
self.clashscore = clashscore(pdb_hierarchy=pdb_hierarchy).get_clashscore()
self.mpscore = molprobity_score(
clashscore = self.clashscore,
rota_out = self.rotamer_outliers,
rama_fav = self.ramachandran_favored)
self.omglz = omegalyze.omegalyze(
pdb_hierarchy=self.pdb_hierarchy, quiet=True)
self.n_cis_proline = self.omglz.n_cis_proline()
self.n_cis_general = self.omglz.n_cis_general()
self.n_twisted_proline = self.omglz.n_twisted_proline()
self.n_twisted_general = self.omglz.n_twisted_general()
def __init__(
self,
pdb_hierarchy,
molprobity_scores=False,
):
self.pdb_hierarchy = pdb_hierarchy
self.clashscore = None
self.ramachandran_outliers = None
self.ramachandran_allowed = None
self.ramachandran_favored = None
self.rotamer_outliers = None
self.c_beta_dev = None
self.mpscore = None
self.omglz = None
self.n_cis_proline = None
self.n_cis_general = None
self.n_twisted_proline = None
self.n_twisted_general = None
if(molprobity_scores):
self.ramalyze_obj = ramalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.ramachandran_outliers = self.ramalyze_obj.percent_outliers
self.ramachandran_allowed = self.ramalyze_obj.percent_allowed
self.ramachandran_favored = self.ramalyze_obj.percent_favored
self.rotalyze_obj = rotalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.rotamer_outliers = self.rotalyze_obj.percent_outliers
self.cbetadev_obj = cbetadev(
pdb_hierarchy = pdb_hierarchy,
outliers_only = True,
out = null_out())
self.c_beta_dev = self.cbetadev_obj.get_outlier_count()
self.clashscore = clashscore(pdb_hierarchy=pdb_hierarchy).get_clashscore()
self.mpscore = molprobity_score(
clashscore = self.clashscore,
rota_out = self.rotamer_outliers,
rama_fav = self.ramachandran_favored)
self.omglz = omegalyze.omegalyze(
pdb_hierarchy=self.pdb_hierarchy, quiet=True)
self.n_cis_proline = self.omglz.n_cis_proline()
self.n_cis_general = self.omglz.n_cis_general()
self.n_twisted_proline = self.omglz.n_twisted_proline()
self.n_twisted_general = self.omglz.n_twisted_general()
def mp_score(self):
return molprobity_score(
clashscore = self.clash().score,
rota_out = self.rotamer().outliers,
rama_fav = self.ramachandran().favored)
def get_model_stat(file_name):
pdb_inp = iotbx.pdb.input(file_name=file_name)
atoms = pdb_inp.atoms()
box = uctbx.non_crystallographic_unit_cell_with_the_sites_in_its_center(
sites_cart=atoms.extract_xyz(), buffer_layer=5)
atoms.set_xyz(new_xyz=box.sites_cart)
ph = pdb_inp.construct_hierarchy()
if (all_single_atom_residues(ph=ph)): return None
raw_recs = ph.as_pdb_string(
crystal_symmetry=box.crystal_symmetry()).splitlines()
#
params = monomer_library.pdb_interpretation.master_params.extract()
params.clash_guard.nonbonded_distance_threshold = None
params.disable_uc_volume_vs_n_atoms_check = False
params.use_neutron_distances = True
params.restraints_library.cdl = False
processed_pdb_file = monomer_library.pdb_interpretation.process(
mon_lib_srv=mon_lib_srv,
ener_lib=ener_lib,
raw_records=raw_recs,
params=params,
log=null_out())
xrs = processed_pdb_file.xray_structure()
sctr_keys = xrs.scattering_type_registry().type_count_dict().keys()
has_hd = "H" in sctr_keys or "D" in sctr_keys
restraints_manager = processed_pdb_file.geometry_restraints_manager(
show_energies=False,
assume_hydrogens_all_missing=not has_hd,
plain_pairs_radius=5.0)
a_mean, b_mean = get_bonds_angles_rmsd(
restraints_manager=restraints_manager, xrs=xrs)
energies_sites = \
restraints_manager.energies_sites(
sites_cart = xrs.sites_cart(),
compute_gradients = False)
nonbonded_distances = energies_sites.nonbonded_distances()
number_of_worst_clashes = (nonbonded_distances < 0.5).count(True)
#
ramalyze_obj = ramalyze(pdb_hierarchy=ph, outliers_only=False)
ramachandran_outliers = ramalyze_obj.percent_outliers
rotamer_outliers = rotalyze(pdb_hierarchy=ph,
outliers_only=False).percent_outliers
c_beta_dev = cbetadev(pdb_hierarchy=ph, outliers_only=True,
out=null_out()).get_outlier_count()
omglz = omegalyze.omegalyze(pdb_hierarchy=ph, quiet=True)
n_cis_proline = omglz.n_cis_proline()
n_cis_general = omglz.n_cis_general()
n_twisted_proline = omglz.n_twisted_proline()
n_twisted_general = omglz.n_twisted_general()
#
clsc = clashscore(pdb_hierarchy=ph).get_clashscore()
mpscore = molprobity_score(clashscore=clsc,
rota_out=rotamer_outliers,
rama_fav=ramalyze_obj.percent_favored)
#
occ = atoms.extract_occ()
bs = atoms.extract_b()
#
return group_args(b_mean=b_mean,
a_mean=a_mean,
number_of_worst_clashes=number_of_worst_clashes,
ramachandran_outliers=ramachandran_outliers,
rotamer_outliers=rotamer_outliers,
c_beta_dev=c_beta_dev,
n_cis_proline=n_cis_proline,
n_cis_general=n_cis_general,
n_twisted_proline=n_twisted_proline,
n_twisted_general=n_twisted_general,
o=occ.min_max_mean().as_tuple(),
b=bs.min_max_mean().as_tuple(),
mpscore=mpscore,
clsc=clsc,
n_atoms=atoms.size())
def __init__(
self,
pdb_hierarchy,
molprobity_scores=False,
):
""" This class is being pickled. Try not to introduce huge members, e.g.
self.hierarchy, etc. This is the reason ramalyze_obj, rotalyze_obj etc
are not members of the class (not self.ramalyze_obj). """
self.clashscore = None
self.ramachandran_outliers = None
self.ramachandran_allowed = None
self.ramachandran_favored = None
self.rotamer_outliers = None
self.c_beta_dev = None
self.mpscore = None
self.omglz = None
self.n_cis_proline = None
self.n_cis_general = None
self.n_twisted_proline = None
self.n_twisted_general = None
if(molprobity_scores):
ramalyze_obj = ramalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.ramachandran_outliers = ramalyze_obj.percent_outliers
self.ramachandran_outliers_cf = ramalyze_obj.get_outliers_count_and_fraction()
self.ramachandran_allowed = ramalyze_obj.percent_allowed
self.ramachandran_allowed_cf = ramalyze_obj.get_allowed_count_and_fraction()
self.ramachandran_favored = ramalyze_obj.percent_favored
self.ramachandran_favored_cf = ramalyze_obj.get_favored_count_and_fraction()
rotalyze_obj = rotalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.rotamer_outliers = rotalyze_obj.percent_outliers
self.rotamer_cf = rotalyze_obj.get_outliers_count_and_fraction()
cbetadev_obj = cbetadev(
pdb_hierarchy = pdb_hierarchy,
outliers_only = True,
out = null_out())
self.c_beta_dev = cbetadev_obj.get_outlier_count()
self.c_beta_dev_percent = cbetadev_obj.get_weighted_outlier_percent()
self.clashscore = clashscore(pdb_hierarchy=pdb_hierarchy).get_clashscore()
self.mpscore = molprobity_score(
clashscore = self.clashscore,
rota_out = self.rotamer_outliers,
rama_fav = self.ramachandran_favored)
omglz = omegalyze.omegalyze(
pdb_hierarchy=pdb_hierarchy, quiet=True)
self.n_proline = omglz.n_proline()
self.n_general = omglz.n_general()
self.n_cis_proline = omglz.n_cis_proline()
self.n_cis_general = omglz.n_cis_general()
self.n_twisted_proline = omglz.n_twisted_proline()
self.n_twisted_general = omglz.n_twisted_general()
self.cis_general = 0
self.twisted_general = 0
self.cis_proline = 0
self.twisted_proline = 0
if self.n_proline != 0:
self.cis_proline = self.n_cis_proline*100./self.n_proline
self.twisted_proline = self.n_twisted_proline*100./self.n_proline
if self.n_general != 0:
self.cis_general = self.n_cis_general*100./self.n_general
self.twisted_general = self.n_twisted_general*100./self.n_general
self.cablam_outliers=None
self.cablam_disfavored=None
self.cablam_ca_outliers=None
try:
cablam_results = cablam.cablamalyze(pdb_hierarchy, outliers_only=False,
out=null_out(), quiet=True)
self.cablam_outliers = cablam_results.percent_outliers()
self.cablam_disfavored = cablam_results.percent_disfavored()
self.cablam_ca_outliers = cablam_results.percent_ca_outliers()
except Exception as e:
print "CaBLAM failed with exception:"
print " %s" % str(e)
pass
def __init__(
self,
pdb_hierarchy,
restraints_manager,
molprobity_scores=False,
n_histogram_slots=10,
cdl_restraints=False,
ignore_hydrogens=False, #only used by amber
):
self.cdl_restraints=cdl_restraints
sites_cart = pdb_hierarchy.atoms().extract_xyz()
energies_sites = \
restraints_manager.energies_sites(
sites_cart = sites_cart,
compute_gradients = False)
# molprobity scores
self.clashscore = None
self.ramachandran_outliers = None
self.ramachandran_allowed = None
self.ramachandran_favored = None
self.rotamer_outliers = None
self.c_beta_dev = None
self.mpscore = None
if(molprobity_scores):
self.ramalyze_obj = ramalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.ramachandran_outliers = self.ramalyze_obj.percent_outliers
self.ramachandran_allowed = self.ramalyze_obj.percent_allowed
self.ramachandran_favored = self.ramalyze_obj.percent_favored
self.rotalyze_obj = rotalyze(pdb_hierarchy=pdb_hierarchy, outliers_only=False)
self.rotamer_outliers = self.rotalyze_obj.percent_outliers
self.cbetadev_obj = cbetadev(
pdb_hierarchy = pdb_hierarchy,
outliers_only = True,
out = null_out())
self.c_beta_dev = self.cbetadev_obj.get_outlier_count()
self.clashscore = clashscore(pdb_hierarchy=pdb_hierarchy).get_clashscore()
self.mpscore = molprobity_score(
clashscore = self.clashscore,
rota_out = self.rotamer_outliers,
rama_fav = self.ramachandran_favored)
#
if(hasattr(energies_sites, "geometry")):
esg = energies_sites.geometry
else: esg = energies_sites
self.a = None
self.b = None
if not hasattr(esg, "angle_deviations"): return
if hasattr(esg, "amber"):
amber_parm = restraints_manager.amber_structs.parm
self.a, angle_deltas = esg.angle_deviations(sites_cart, amber_parm,
ignore_hd=ignore_hydrogens,
get_deltas=True)
self.b, bond_deltas = esg.bond_deviations(sites_cart, amber_parm,
ignore_hd=ignore_hydrogens,
get_deltas=True)
self.a_number = esg.n_angle_proxies(amber_parm,
ignore_hd=ignore_hydrogens)
self.b_number = esg.n_bond_proxies(amber_parm,
ignore_hd=ignore_hydrogens)
self.c, self.p, self.ll, self.d, self.n = None, None, None, None, None
self.c_number=0
self.p_number=0
self.d_number=0
self.bond_deltas_histogram = \
flex.histogram(data = flex.abs(bond_deltas), n_slots = n_histogram_slots)
self.angle_deltas_histogram = \
flex.histogram(data = flex.abs(angle_deltas), n_slots = n_histogram_slots)
# nonbonded_distances = esg.nonbonded_distances()
# self.nonbonded_distances_histogram = flex.histogram(
# data = flex.abs(nonbonded_distances), n_slots = n_histogram_slots)
for restraint_type in ["b", "a", "c", "p", "ll", "d", "n"] :
for value_type in [("mean",2), ("max",1), ("min",0)] :
name = "%s_%s" % (restraint_type, value_type[0])
if getattr(self, restraint_type) is None:
setattr(self, name, None)
continue
setattr(self, name, getattr(self, restraint_type)[value_type[1]])
return
self.a = esg.angle_deviations()
self.b = esg.bond_deviations()
self.a_number = esg.get_filtered_n_angle_proxies()
self.b_number = esg.get_filtered_n_bond_proxies()
self.c = esg.chirality_deviations()
self.d = esg.dihedral_deviations()
self.p = esg.planarity_deviations()
self.ll = esg.parallelity_deviations()
self.n = esg.nonbonded_deviations()
self.c_number = esg.n_chirality_proxies
self.d_number = esg.n_dihedral_proxies
self.p_number = esg.n_planarity_proxies
self.n_number = esg.n_nonbonded_proxies
#
for restraint_type in ["b", "a", "c", "p", "ll", "d", "n"] :
for value_type in [("mean",2), ("max",1), ("min",0)] :
name = "%s_%s" % (restraint_type, value_type[0])
if getattr(self, restraint_type) is None: continue
setattr(self, name, getattr(self, restraint_type)[value_type[1]])
#
if(hasattr(restraints_manager, "geometry")):
rmg = restraints_manager.geometry
else: rmg = restraints_manager
bond_deltas = geometry_restraints.bond_deltas(
sites_cart = sites_cart,
sorted_asu_proxies = rmg.pair_proxies().bond_proxies)
angle_deltas = geometry_restraints.angle_deltas(
sites_cart = sites_cart,
proxies = rmg.angle_proxies)
nonbonded_distances = esg.nonbonded_distances()
self.bond_deltas_histogram = \
flex.histogram(data = flex.abs(bond_deltas), n_slots = n_histogram_slots)
self.angle_deltas_histogram = \
flex.histogram(data = flex.abs(angle_deltas), n_slots = n_histogram_slots)
self.nonbonded_distances_histogram = flex.histogram(
data = flex.abs(nonbonded_distances), n_slots = n_histogram_slots)
#
assert approx_equal(
esg.target,
esg.angle_residual_sum+
esg.bond_residual_sum+
esg.chirality_residual_sum+
esg.dihedral_residual_sum+
esg.nonbonded_residual_sum+
esg.planarity_residual_sum+
esg.parallelity_residual_sum+
esg.reference_coordinate_residual_sum+
esg.reference_dihedral_residual_sum+
esg.ncs_dihedral_residual_sum+
esg.den_residual_sum+
esg.ramachandran_residual_sum)
del energies_sites, esg # we accumulate this object, so make it clean asap
