def __init__(self,
pdb_hierarchy,
params,
map_coeffs=None,
difference_map_coeffs=None,
ccp4_map=None,
unit_cell=None,
grid_spacing=0.2,
sampling_method="linear",
n_chi_max=4,
log=None):
if (log is None): log = sys.stdout
adopt_init_args(self, locals())
models = pdb_hierarchy.models()
if (len(models) > 1):
raise Sorry("Multi-model PDB files not supported.")
self.sigma = self.real_map = self.difference_map = None
if (map_coeffs is not None):
self.unit_cell = map_coeffs.unit_cell()
if (params.sampling_method == "direct"):
self.map_coeffs = self.map_coeffs.expand_to_p1()
if (not map_coeffs.anomalous_flag()):
self.map_coeffs = self.map_coeffs.generate_bijvoet_mates()
if (sampling_method != "direct") or (params.scaling == "sigma"):
fft_map = self.map_coeffs.fft_map(
resolution_factor=grid_spacing)
if (params.scaling == "sigma"):
self.sigma = fft_map.statistics().sigma()
fft_map.apply_sigma_scaling()
else:
fft_map.apply_volume_scaling()
self.real_map = fft_map.real_map_unpadded()
else:
assert (ccp4_map is not None)
print >> self.log, "CCP4 map statistics:"
ccp4_map.show_summary(out=self.log, prefix=" ")
self.real_map = ccp4_map.data.as_double()
# XXX assume that the map is already scaled properly (in the original
# unit cell)
self.sigma = 1 #ccp4_map.statistics().sigma()
# XXX the unit cell that we need for the non-crystallographic
# interpolation is not what comes out of the map - it's the
self.unit_cell = ccp4_map.grid_unit_cell()
if (difference_map_coeffs is not None):
if (sampling_method == "direct"):
self.difference_map_coeffs = self.difference_map_coeffs.expand_to_p1(
)
if (not difference_map_coeffs.anomalous_flag()):
self.difference_map_coeffs = \
self.difference_map_coeffs.generate_bijvoet_mates()
if (sampling_method != "direct") or (params.scaling == "sigma"):
fft_map = self.difference_map_coeffs.fft_map(
resolution_factor=params.grid_spacing)
if (params.scaling == "sigma"):
fft_map.apply_sigma_scaling()
else:
fft_map.apply_volume_scaling()
self.difference_map = fft_map.real_map_unpadded()
results = []
from mmtbx.rotamer import sidechain_angles
self.angle_lookup = sidechain_angles.SidechainAngles(False)
self.sites_cart = pdb_hierarchy.atoms().extract_xyz()
self.residue_groups = []
for chain in models[0].chains():
self.residue_groups.extend(chain.residue_groups())
if (params.nproc in [None, Auto]) or (params.nproc > 1):
# this will be a list of lists
results_ = easy_mp.pool_map(processes=params.nproc,
fixed_func=self.__sample_density,
args=range(len(self.residue_groups)))
# now flatten it out
self.results = []
for result_list in results_:
self.results.extend(result_list)
else:
self.results = []
for i_res in range(len(self.residue_groups)):
self.results.extend(self.__sample_density(i_res, verbose=True))
# correlations between residues in one ensemble a la MutInf...
#
# Started by Daniel Keedy in May 2013
#
##############################################################
import sys
import os
import iotbx.pdb.hierarchy
from mmtbx.rotamer import rotamer_eval
from mmtbx.rotamer import sidechain_angles
import iotbx.pdb.amino_acid_codes
# These scorers are global variables since they hog memory (I think?)
rotamer_scorer = rotamer_eval.RotamerEval()
sc_angles_measurer = sidechain_angles.SidechainAngles(show_errs=True)
aa_resnames = iotbx.pdb.amino_acid_codes.one_letter_given_three_letter
class StructureEnsemble():
def __init__(self, nm):
self.name = nm
self.conf_ensems = {} # CNIT --> ConformerEnsemble
def add_structures(self, dirname):
for filename in os.listdir(dirname):
self.add_structure(filename)
def add_structure(self, filename):
pdb = iotbx.pdb.hierarchy.input(file_name=filename)
def __init__ (self,
pdb_hierarchy,
params,
map_coeffs=None,
crystal_symmetry_model=None,
difference_map_coeffs=None,
ccp4_map=None,
unit_cell=None,
grid_spacing=0.2,
sampling_method="linear",
n_chi_max=4,
log=None) :
if (log is None) : log = sys.stdout
adopt_init_args(self, locals())
models = pdb_hierarchy.models()
if (len(models) > 1) :
raise Sorry("Multi-model PDB files not supported.")
self.sigma = self.real_map = self.difference_map = None
if (map_coeffs is not None) :
self.unit_cell = map_coeffs.unit_cell()
if (params.sampling_method == "direct") :
self.map_coeffs = self.map_coeffs.expand_to_p1()
if (not map_coeffs.anomalous_flag()) :
self.map_coeffs = self.map_coeffs.generate_bijvoet_mates()
if (sampling_method != "direct") or (params.scaling == "sigma") :
fft_map = self.map_coeffs.fft_map(resolution_factor=grid_spacing)
if (params.scaling == "sigma") :
self.sigma = fft_map.statistics().sigma()
fft_map.apply_sigma_scaling()
else :
fft_map.apply_volume_scaling()
self.real_map = fft_map.real_map_unpadded()
else :
assert (ccp4_map is not None)
print >> self.log, "CCP4 map statistics:"
ccp4_map.show_summary(out=self.log, prefix=" ")
space_group_number = ccp4_map.space_group_number
from cctbx import crystal
crystal_symmetry_map = crystal.symmetry(ccp4_map.unit_cell().parameters(), space_group_number)
if not crystal_symmetry_model:
print >> self.log, """Warning: the model does not contain symmetry information. Using map information."""
elif not crystal_symmetry_map.is_similar_symmetry(crystal_symmetry_model):
print >> self.log, """Warning: The map and model appear to have different crystal symmetry information.
EMRinger will assume the map symmetry data is correct and process."""
# If map space group is P1, then check that model space group is also either not present or is P1.
# If both are p1 or model symmetry is not present, then do the shift. Otherwise, no shift.
if space_group_number == 1 and not (crystal_symmetry_model and crystal_symmetry_model.space_group() and crystal_symmetry_model.space_group().info().type().number() != 1):
import mmtbx.utils
shift_manager = mmtbx.utils.shift_origin(
map_data = ccp4_map.data.as_double(),
pdb_hierarchy = pdb_hierarchy,
crystal_symmetry = crystal_symmetry_map)
if not shift_manager.shift_cart == None:
print >> self.log, "Warning: Model and Map use different origin. Applying origin shift to compensate."
pdb_hierarchy = shift_manager.pdb_hierarchy # gives you shifted model
self.real_map = shift_manager.map_data # gives you shifted map
else:
print >> self.log, """Warning: Structure is not P1, so automatic origin shifts cannot currently be applied"""
self.real_map = ccp4_map.data.as_double()
# XXX assume that the map is already scaled properly (in the original
# unit cell)
models = pdb_hierarchy.models()
self.sigma = 1 #ccp4_map.statistics().sigma()
# XXX the unit cell that we need for the non-crystallographic
# interpolation is not what comes out of the map - it's the
self.unit_cell = ccp4_map.grid_unit_cell()
if (difference_map_coeffs is not None) :
if (sampling_method == "direct") :
self.difference_map_coeffs = self.difference_map_coeffs.expand_to_p1()
if (not difference_map_coeffs.anomalous_flag()) :
self.difference_map_coeffs = \
self.difference_map_coeffs.generate_bijvoet_mates()
if (sampling_method != "direct") or (params.scaling == "sigma") :
fft_map = self.difference_map_coeffs.fft_map(
resolution_factor=params.grid_spacing)
if (params.scaling == "sigma") :
fft_map.apply_sigma_scaling()
else :
fft_map.apply_volume_scaling()
self.difference_map = fft_map.real_map_unpadded()
results = []
from mmtbx.rotamer import sidechain_angles
self.angle_lookup = sidechain_angles.SidechainAngles(False)
self.sites_cart = pdb_hierarchy.atoms().extract_xyz()
self.residue_groups = []
for chain in models[0].chains() :
self.residue_groups.extend(chain.residue_groups())
if (params.nproc in [None,Auto]) or (params.nproc > 1) :
# this will be a list of lists
results_ = easy_mp.pool_map(
processes=params.nproc,
fixed_func=self.__sample_density,
args=range(len(self.residue_groups)))
# now flatten it out
self.results = []
for result_list in results_ : self.results.extend(result_list)
else :
self.results = []
for i_res in range(len(self.residue_groups)) :
self.results.extend(self.__sample_density(i_res, verbose=True))
if len(self.results) == 0:
raise Sorry("""No residues could be scanned by EMRinger, so scores cannot be generated.
There are a few problems that can lead to this, including not having
modeled side chains (poly-A or poly-G models), mismatches between the map
and model grid, or corrupted map density values. These problems can often
be assessed with molecular graphics tools such as pymol or coot.""")
