def simple_occupancy_groups(hierarchy,
include_single_conformer_groups=False,
verbose=False):
"""Given a selection, return the default occupancy groups"""
occupancy_groups = []
# Iterate through the default occupancy groups
for g in hierarchy.occupancy_groups_simple():
# Skip single groups
if (len(g) == 1) and (len(g[0]) == 1):
if verbose:
print 'Not making simple restraints for single-atom groups:', ','.join(
[
Labeller.format(a)
for a in hierarchy.select(flex.size_t(g[0])).atoms()
])
continue
if (len(g) == 1) and (not include_single_conformer_groups):
if verbose:
print 'Not making simple restraints for single-conformer groups:\n\t', '\n\t'.join(
[
Labeller.format(a) for a in hierarchy.select(
flex.size_t(g[0])).atom_groups()
])
continue
selections = []
for sel in g:
ags = [
GenericSelection.to_dict(ag)
for ag in hierarchy.select(flex.size_t(sel)).atom_groups()
]
selections.append(ags)
occupancy_groups.append(selections)
return occupancy_groups
def expand_alternate_conformations(hierarchy, in_place=False, verbose=False):
"""Convert all atoms to multiple conformers - full multi-conformer representation of the model"""
if not in_place: hierarchy = hierarchy.deep_copy()
# Get all of the altlocs that should be present for each atom
full_altloc_set = sorted([a for a in hierarchy.altloc_indices() if a])
# If not altlocs found, expand all to "A"
if full_altloc_set == []:
if verbose:
print 'No altlocs in structure: expanding all residues to conformer "A"'
full_altloc_set = ['A']
if verbose:
print 'Expanding all (appropriate) residues to have altlocs {}'.format(
full_altloc_set)
print '------------------>'
# Iterate through and expand each residue group to have all conformers
for chain in hierarchy.chains():
for residue_group in chain.residue_groups():
# If has conformers but has blank altloc atoms (add blank ag to all other ags)
if residue_group.have_conformers(
) and residue_group.move_blank_altloc_atom_groups_to_front():
if verbose:
print '{} - expanding to pure conformer (altlocs {})'.format(
Labeller.format(residue_group),
[a.altloc for a in residue_group.atom_groups()])
# Convert all residue_groups to pure alt-conf
create_pure_alt_conf_from_proper_alt_conf(
residue_group=residue_group, in_place=True)
# Can go to next if all conformers are present for this residue group
current_set = {a.altloc for a in residue_group.atom_groups()}
if not current_set.symmetric_difference(full_altloc_set): continue
# Only want to expand conformers for protein atoms (which should be present in all conformers)
# or where the residue group is only present in one conformation (single conformer water)
# but DO NOT want to expand waters in conformer A to A,B,C etc...
if protein_amino_acid_set.intersection(
residue_group.unique_resnames()) or (
not residue_group.have_conformers()):
if verbose:
print '{} - populating missing conformers (current altlocs {}, target set {})'.format(
Labeller.format(residue_group), current_set,
full_altloc_set)
# Populate missing conformers (from the other conformers)
populate_missing_conformers(residue_group=residue_group,
full_altloc_set=full_altloc_set,
in_place=True,
verbose=verbose)
assert sorted([a.altloc for a in residue_group.atom_groups()
]) == full_altloc_set
if verbose:
print '{} - updated conformer list: (current altlocs {}, target set {})'.format(
Labeller.format(residue_group),
[a.altloc for a in residue_group.atom_groups()],
full_altloc_set)
if verbose: print '------------------>'
return hierarchy
def set_conformer_occupancy(hierarchy, altlocs, occupancy, in_place=False, verbose=False):
"""Normalise the occupancies of a hierarchy so that the occupancies for a residue sum to 1.0"""
if isinstance(altlocs, str): altlocs=list(altlocs)
else: assert isinstance(altlocs, list), 'altlocs must be either str or list'
if (not in_place): hierarchy = hierarchy.deep_copy()
for ag in hierarchy.atom_groups():
if ag.altloc in altlocs:
if verbose: print '{} - setting occupancy to {}'.format(Labeller.format(ag), occupancy)
ag.atoms().set_occ(flex.double(ag.atoms().size(), occupancy))
return hierarchy
def increment_altlocs(hierarchy, offset=1, in_place=False, verbose=False):
"""Increment all altlocs in the structure by a certain number of letters"""
if not in_place: hierarchy = hierarchy.deep_copy()
cur_altlocs = [a for a in hierarchy.altloc_indices() if a]
all_altlocs = iotbx.pdb.systematic_chain_ids()
new_altlocs = dict([(a, all_altlocs[all_altlocs.index(a) + offset])
for a in cur_altlocs])
if verbose:
print '------------------>'
print 'Updating altlocs:'
for a in cur_altlocs:
print '{} -> {}'.format(a, new_altlocs[a])
print '------------------>'
for atom_group in hierarchy.atom_groups():
if verbose:
print '{} - updating altloc: {} -> {}'.format(
Labeller.format(atom_group), atom_group.altloc,
new_altlocs[atom_group.altloc])
assert atom_group.altloc != ''
atom_group.altloc = new_altlocs[atom_group.altloc]
return hierarchy
def create_levels_tab(parameterisation):
p = parameterisation
f = parameterisation.fitter
fm = parameterisation.file_manager
chain_ids = [c.id for c in p.blank_master_hierarchy().select(flex.bool(p.atom_mask.tolist()),copy_atoms=True).chains()]
tab = {'id' : 'levels',
'short_name' : 'ADP Summary',
'long_name' : 'Level-by-level TLS parameterisation',
'description': 'Parameteristaion composed of {} levels.'.format(len(f.levels)),
'tabs' : [],
}
# -------------------------------->
# Create overview sub-tab
# -------------------------------->
overview_tab = {'id' : tab['id']+'overview',
'active' : True,
'short_name' : 'Overview',
'long_name' : 'Overview of the parameterised hierarchical ADP model',
'description' : '',
'panels' : [],
}
tab['tabs'].append(overview_tab)
# Split the panels up by chain
for c_id in chain_ids:
# Split up the chains with divider panels
prof_f = fm.get_file('png-combined-profile-template').format(c_id)
resd_f = fm.get_file('png-residual-profile-template').format(c_id)
panel = {'id' : '<h4>Levels for Chain {}</h4>'.format(c_id),
'width' : 12,
'show' : True,
'table' : None,
'objects' : [{'width':6, 'text': 'TLS-level components', 'path':png2base64src_maybe(prof_f, print_on_missing=DEBUG)},
{'width':6, 'text': 'Residual component', 'path':png2base64src_maybe(resd_f, print_on_missing=DEBUG)}],
}
overview_tab['panels'].append(panel)
# Add images to the overview tab for each TLS level
for i_level, (level_num, level_lab, level) in enumerate(f):
chain_image = fm.get_file('pml-level-chain-template').format(level_num, c_id)
stack_image = fm.get_file('png-tls-profile-template').format(level_num, c_id)
aniso_image = fm.get_file('png-tls-anisotropy-template').format(level_num, c_id)
panel = {'id' : 'Level {} of {} ({})'.format(level_num, len(f.levels),level_lab),
'width' : 4,
'show' : True,
'table' : None,
'objects' : [{'width':12, 'text':'{} atoms.'.format('X')},
{'width':12, 'path':png2base64src_maybe(chain_image, print_on_missing=DEBUG)},
{'width':12, 'path':png2base64src_maybe(stack_image, print_on_missing=DEBUG)},
{'width':12, 'path':png2base64src_maybe(aniso_image, print_on_missing=DEBUG)}],
}
overview_tab['panels'].append(panel)
# Format residual level
chain_image = fm.get_file('pml-residual-chain-template').format(c_id)
stack_image = fm.get_file('png-residual-profile-template').format(c_id)
aniso_image = fm.get_file('png-residual-anisotropy-template').format(c_id)
panel = {'id' : 'Final Level (residual)',
'width' : 4,
'show' : True,
'table' : None,
'objects' : [{'width':12, 'text':'{} atoms.'.format('X')},
{'width':12, 'path':png2base64src_maybe(chain_image, print_on_missing=DEBUG)},
{'width':12, 'path':png2base64src_maybe(stack_image, print_on_missing=DEBUG)},
{'width':12, 'path':png2base64src_maybe(aniso_image, print_on_missing=DEBUG)}],
}
overview_tab['panels'].append(panel)
# -------------------------------->
# Create tab for each level
# -------------------------------->
for i_level, (level_num, level_lab, level) in enumerate(f):
# Create dictionary for this tab and add to tab_list
level_tab = {'id' : tab['id']+'lvl{}'.format(level_num),
'short_name' : 'Level {}'.format(level_num),
'long_name' : 'Level {} ({})'.format(level_num, level_lab),
'description': 'Level {} of {}. '.format(level_num, len(f.levels))+\
'Composed of {} groups'.format(level.n_groups()),
'panels' : [],
}
tab['tabs'].append(level_tab)
# Add overview at the top of the tab
for c_id in chain_ids:
partn_image = fm.get_file('pml-level-partition-template').format(level_num, c_id)
chain_image = fm.get_file('pml-level-chain-template').format(level_num, c_id)
stack_image = fm.get_file('png-tls-profile-template').format(level_num, c_id)
aniso_image = fm.get_file('png-tls-anisotropy-template').format(level_num, c_id)
panel = {'id' : 'Chain {}'.format(c_id),
'width' : 12,
'show' : True,
'table' : None,
'objects' : [{'width':12, 'text':'{} atoms.'.format('X')},
{'width':6, 'path':png2base64src_maybe(partn_image, print_on_missing=DEBUG)},
{'width':6, 'path':png2base64src_maybe(chain_image, print_on_missing=DEBUG)},
{'width':6, 'path':png2base64src_maybe(stack_image, print_on_missing=DEBUG)},
{'width':6, 'path':png2base64src_maybe(aniso_image, print_on_missing=DEBUG)}],
}
level_tab['panels'].append(panel)
# Read in the TLS models and amplitudes for this level
tls_models = pandas.read_csv(fm.get_file('csv-tls-mdl-template').format(level_num)).set_index(['group','model']).drop('Unnamed: 0', axis=1, errors='ignore')
tls_amplitudes = pandas.read_csv(fm.get_file('csv-tls-amp-template').format(level_num)).set_index(['group','model','cpt']).drop('Unnamed: 0', axis=1, errors='ignore')
# Extract groups for each level
for i_group, (group_num, sel, group_fitter) in enumerate(level):
# Extract TLS values for this group
tls_vals = [tls_models.loc[(group_num, i_mode)] for i_mode in xrange(p.params.fitting.tls_models_per_tls_group)]
# Skip if no TLS values
if numpy.abs(tls_vals).sum() == 0.0:
continue
# Get images and format values
scl_image = fm.get_file('pml-level-scaled-template').format(level_num, group_num)
adp_image = fm.get_file('pml-level-group-template').format(level_num, group_num)
amp_image = fm.get_file('png-tls-amp-dist-template').format(level_num, group_num)
tls_mdl_strs = [('Mode {}:<br>' + \
'<samp>\n' + \
'T: {T11:>9.3f}, {T22:>9.3f}, {T33:>9.3f}, {T12:>9.3f}, {T13:>9.3f}, {T23:>9.3f},<br>' + \
'L: {L11:>9.3f}, {L22:>9.3f}, {L33:>9.3f}, {L12:>9.3f}, {L13:>9.3f}, {L23:>9.3f},<br>' + \
'S: {S11:>9.3f}, {S12:>9.3f}, {S13:>9.3f}, {S21:>9.3f}, {S22:>9.3f}, {S23:>9.3f}, {S31:>9.3f}, {S32:>9.3f}, {S33:>9.3f}' + \
'\n</samp>'
).format(i_mode+1, **mode_vals.round(3)).replace(' ',' ') if mode_vals.any() else 'Zero-value TLS values for mode {}'.format(i_mode+1) for i_mode, mode_vals in enumerate(tls_vals)]
# Create panel dictionary
panel = {'id' : 'Group {} - {}'.format(group_num, p.levels[i_level][i_group]),
'width' : 12, #max(4,12//level.n_groups()),
'table' : None,
'objects': [{'width':12, 'text':'<br>'.join(['Number of atoms: {}'.format(sum(sel))])},
{'width':4, 'text':'Shape of disorder (arbitrary scale)', 'path': png2base64src_maybe(scl_image, print_on_missing=DEBUG)},
{'width':4, 'text':'Average size over all datasets', 'path': png2base64src_maybe(adp_image, print_on_missing=DEBUG)},
{'width':4, 'text':'Amplitude Distribution', 'path': png2base64src_maybe(amp_image, print_on_missing=DEBUG)}] + \
[{'width':12,'text':s} for s in tls_mdl_strs],
}
level_tab['panels'].append(panel)
# Make the first panel open
if len(level_tab['panels']) > 0:
level_tab['panels'][0]['show'] = True
# -------------------------------->
# Create tab for residual level
# -------------------------------->
residual_tab = {'id' : 'lvlres',
'short_name' : 'Residual',
'long_name' : 'Final Level (residual)',
'description': '',
'panels' : [],
}
tab['tabs'].append(residual_tab)
# Get selection for fitted atoms
atom_sel = flex.bool(p.atom_mask.tolist())
# Create row for each residue
for i_chain, c in enumerate(p.blank_master_hierarchy().select(atom_sel,copy_atoms=True).chains()):
panel = {'id' : 'Residual components for chain {}'.format(c.id),
'width' : 12,
'table' : None,
'objects': [],
}
residual_tab['panels'].append(panel)
for i_rg, rg in enumerate(c.residue_groups()):
short_label = ShortLabeller.format(rg)
long_label = Labeller.format(rg)
adp_image = fm.get_file('pml-residual-group-template').format(short_label)
panel['objects'].append({'width':4, 'text':long_label, 'path': png2base64src_maybe(adp_image, print_on_missing=DEBUG)})
# Make the first panel open
residual_tab['panels'][0]['show'] = True
return tab
def transfer_residue_groups_from_other(acceptor_hierarchy,
donor_hierarchy,
in_place=False,
verbose=False):
"""Transfer atom_groups from donor_hierarchy to matching residue_groups in acceptor_hierarchy, creating new chains and residue groups only where necessary"""
if not in_place: acceptor_hierarchy = acceptor_hierarchy.deep_copy()
# Sort all residues (by chain then id) for the acceptor hierarchy
accept_model = acceptor_hierarchy.only_model()
accept_dict = {c.id: {} for c in accept_model.chains()}
[
accept_dict.get(rg.parent().id).setdefault(rg.resid(), []).append(rg)
for rg in accept_model.residue_groups()
]
# Dictionary to link matching chains (allows multiple chain As to be linked uniquely to multiple chain As)
link_dict = {}
# Residues that don't have a matching partner in the old hierarchy
tricky_rgs = []
# Iterate through donor chains
for donor_ch in donor_hierarchy.only_model().chains():
# If chain not in hierarchy, simply copy across
if accept_dict.get(donor_ch.id, None) is None:
if verbose:
print 'Transferring whole chain: {}'.format(
Labeller.format(donor_ch))
accept_model.append_chain(donor_ch.detached_copy())
continue
# Chain present, copy by residue_group
for donor_rg in donor_ch.residue_groups():
# Find equivalent residue groups in the other hierarchy
accept_rg = accept_dict.get(donor_ch.id).get(donor_rg.resid(), [])
if len(accept_rg) > 1:
# Should only be one...
raise Exception(
'More than one residue group in hierarchy with the same residue_id and chain_id'
)
elif len(accept_rg) == 1:
accept_rg = accept_rg[0]
# Record the links between these chains
link_dict.setdefault(donor_rg.parent(), accept_rg.parent())
# Transfer atom groups to this residue_group
if verbose:
print 'Transferring atom groups: {} > {}'.format(
Labeller.format(donor_rg), Labeller.format(accept_rg))
for donor_ag in donor_rg.atom_groups():
accept_rg.append_atom_group(donor_ag.detached_copy())
else:
# Have the possibility of multiple chains with the same id, so at the moment, store for later
tricky_rgs.append(donor_rg)
# Transfer residues that have chain matches, but don't have residue matches in the acceptor structures
for donor_rg in tricky_rgs:
# Get chain from link_dict
accept_ch = link_dict.get(donor_rg.parent().id, None)
# If the chain isn't linked:
if accept_ch is None:
# If there's only one chain with the same ID, choose this one
possible_chains = [
c for c in accept_model.chains()
if c.id == donor_rg.parent().id
]
if len(possible_chains) == 1:
accept_ch = possible_chains[0]
else:
raise Failure(
"Don't know how to transfer {} to the output model".format(
Labeller.format(donor_rg)))
# Simply append to chain
if verbose:
print 'Transferring residue group: {} > {}'.format(
Labeller.format(donor_rg), Labeller.format(accept_ch))
accept_ch.append_residue_group(donor_rg.detached_copy())
return acceptor_hierarchy
def resolve_residue_id_clashes(fixed_hierarchy,
moving_hierarchy,
in_place=False,
verbose=False):
"""Move residues in mov_hierarchiy to new chains if they have the same resid as a residue in fixed_hierarchy but different resnames"""
if (not in_place): moving_hierarchy = moving_hierarchy.deep_copy()
# Sort all residues (by chain then id) for the fixed hierarchy - chains > residue_ids > residue_groups (objects)
fixed_dict = {c.id: {} for c in fixed_hierarchy.only_model().chains()}
[
fixed_dict.get(rg.parent().id).setdefault(rg.resid(), []).append(rg)
for rg in fixed_hierarchy.residue_groups()
]
# Find the residues with clashing resids
residues_to_update = []
for rg_mov in moving_hierarchy.only_model().residue_groups():
# Extract equivalent group for this residue
rg_ref = fixed_dict.get(rg_mov.parent().id, {}).get(rg_mov.resid(), [])
# Check that we've found only one (rather than silently erroring)
if len(rg_ref) == 0:
continue
elif len(rg_ref) > 1:
raise Exception(
'More than one matching residue group in the output hierarchy?'
)
rg_ref = rg_ref[0]
# Check to see if the residue is the same type as in the reference structure
if map(str.strip,
rg_ref.unique_resnames()) == map(str.strip,
rg_mov.unique_resnames()):
# Same residue -- that's fine
# TODO allow this if the resnames match OR if it's a protein residue? (allows mutations/reactions?)
continue
else:
# Will need to be added to different chain
if verbose:
print 'Different residues with same id - changing chains: {} != {}'.format(
list(rg_ref.unique_resnames()),
list(rg_mov.unique_resnames()))
residues_to_update.append(rg_mov)
# Nothing to do -- return
if not residues_to_update:
return moving_hierarchy
# New chain to add clashing groups to
new_chain = None
old_chain_prev = None
new_chain_ids = find_unused_chain_ids(
hierarchies=[fixed_hierarchy, moving_hierarchy])
# Go through and transfer the residue groups to new chains
for rg_mov in residues_to_update:
old_chain = rg_mov.parent()
# See if there is a residue with this id already present in the new chain
if (not new_chain) or (rg_mov.resid() in new_chain.get_residue_ids()
) or (old_chain_prev.id != old_chain.id):
new_chain = iotbx.pdb.hierarchy.chain(id=new_chain_ids.pop(0))
old_chain.parent().append_chain(new_chain)
if verbose:
print '{} - moving to chain {}'.format(Labeller.format(rg_mov),
new_chain.id)
# Remove from old chain and add to the new chain
old_chain.remove_residue_group(rg_mov)
new_chain.append_residue_group(rg_mov)
# Keep track of old_chain of previous residue
old_chain_prev = old_chain
return moving_hierarchy
def sanitise_occupancies(hierarchy,
fixed_conformers=None,
min_occ=0.0,
max_occ=1.0,
in_place=False,
verbose=False):
"""Sanitise the occupancies of a hierarchy so that the occupancies for a residue sum to 1.0"""
assert (min_occ >= 0.0) and (max_occ <= 1.0)
if fixed_conformers is None: fixed_conformers = []
if (not in_place): hierarchy = hierarchy.deep_copy()
# Iterate through the output structure, and normalise the occupancies if necessary
for rg in hierarchy.residue_groups():
# Calculate occupancy of the residue group
rg_occ = calculate_residue_group_occupancy(residue_group=rg)
# If occupancy in range, continue
if min_occ <= rg_occ <= max_occ:
continue
if verbose:
print 'Occupancy of residue {} is {} -- sanitising'.format(
Labeller.format(rg), rg_occ)
# Extract main-conf and alt-conf ags
main_ag, alt_ags = split_main_and_alt_conf_atom_groups(rg)
# Sanitise main conf
if main_ag is not None:
if verbose:
print '------------------>'
print 'Sanitising main-conf atom group:\n\t{}'.format(
Labeller.format(main_ag))
print 'Current occupancy: {}'.format(
max(main_ag.atoms().extract_occ()))
sanitise_atom_group_occupancies_in_place(main_ag,
min_occ=min_occ,
max_occ=max_occ)
if verbose:
print 'New occupancy: {}'.format(
max(main_ag.atoms().extract_occ()))
# Sanitise alt confs
if alt_ags is not None:
# Get the groups to change and the groups to keep constant
ag_chnge = [
ag for ag in alt_ags if (ag.altloc not in fixed_conformers)
]
ag_const = [
ag for ag in alt_ags if (ag.altloc in fixed_conformers)
]
# Calculate the total occupancy of the groups
occ_chnge = sum([max(ag.atoms().extract_occ()) for ag in ag_chnge])
occ_const = sum([max(ag.atoms().extract_occ()) for ag in ag_const])
if occ_const > max_occ:
raise Exception(
'Occupancy of fixed atom groups ({}) is already greater than maximum ({})'
.format(occ_const, max_occ))
# Normalise the occupancies of the changing groups
if verbose:
print '------------------>'
print 'Sanitising alt-conf atom groups:'
print 'Fixed conformers:\n\t{}'.format('\n\t'.join(
[Labeller.format(ag) for ag in ag_const]))
print 'Other conformers:\n\t{}'.format('\n\t'.join(
[Labeller.format(ag) for ag in ag_chnge]))
print 'Total occupancy (fixed): {}'.format(occ_const)
print 'Individual occupancies: {}'.format(', '.join(
map(str,
[max(ag.atoms().extract_occ()) for ag in ag_const])))
print 'Total occupancy (other): {}'.format(occ_chnge)
print 'Individual occupancies: {}'.format(', '.join(
map(str,
[max(ag.atoms().extract_occ()) for ag in ag_chnge])))
sanitise_multiple_atom_group_occupancies_in_place(
atom_groups=ag_chnge,
min_occ=max(0.0, min_occ - occ_const),
max_occ=min(1.0, max_occ - occ_const))
if verbose:
print 'New total occupancy (other): {}'.format(
sum([max(ag.atoms().extract_occ()) for ag in ag_chnge]))
print 'Individual occupancies: {}'.format(', '.join(
map(str,
[max(ag.atoms().extract_occ()) for ag in ag_chnge])))
return hierarchy
def prune_redundant_alternate_conformations(hierarchy,
required_altlocs=[],
rmsd_cutoff=0.1,
in_place=False,
verbose=False):
"""Remove alternate conformers of residues if residues has conformers of required_altlocs and all conformers are within rmsd_cutoff"""
if not in_place: hierarchy = hierarchy.deep_copy()
required_altlocs = set(required_altlocs)
for residue_group in hierarchy.residue_groups():
# Skip if no conformers
if not residue_group.have_conformers():
continue
# Get the blank and non-blank altloc atom_groups
if residue_group.move_blank_altloc_atom_groups_to_front() != 0:
main_ag = residue_group.atom_groups()[0]
alt_ags = residue_group.atom_groups()[1:]
assert main_ag.altloc == ''
assert alt_ags != []
else:
main_ag = None
alt_ags = residue_group.atom_groups()
# Check no misplaced main conf
assert '' not in [ag.altloc for ag in alt_ags]
# Check if required altlocs are present (return if not)
if required_altlocs.difference([ag.altloc for ag in alt_ags]):
continue
# Check if all pair of conformers are within rmsd cutoff
prune = True
for i, ag_1 in enumerate(alt_ags):
for j, ag_2 in enumerate(alt_ags):
if j <= i: continue
d = calculate_paired_atom_rmsd(atoms_1=ag_1.atoms(),
atoms_2=ag_2.atoms(),
sort=True,
truncate_to_common_set=False)
if verbose:
print 'Residue {}, alt {} - alt {}: rmsd {}'.format(
Labeller.format(residue_group), i, j, d)
if (d is None) or (d > rmsd_cutoff):
prune = False
break
if prune is False: break
if prune is False: continue
# All rmsds below cutoff - prune!
if verbose:
print 'Pruning {}: altlocs {} -> [""]'.format(
Labeller.format(residue_group), [ag.altloc for ag in alt_ags])
if main_ag:
# Merge one alt group with the main atom_group
new_main_ag = alt_ags[0].detached_copy()
new_main_ag.altloc = ''
normalise_occupancies(atoms=new_main_ag.atoms(),
max_occ=max(main_ag.atoms().extract_occ()))
residue_group.merge_atom_groups(main_ag, new_main_ag)
else:
# Remove one atom_group and set altloc to ''
new_main_ag = alt_ags.pop(0)
new_main_ag.altloc = ''
normalise_occupancies(atoms=new_main_ag.atoms(),
max_occ=sum([
max(ag.atoms().extract_occ())
for ag in [new_main_ag] + alt_ags
]))
# Remove all remaining alternate groups
[residue_group.remove_atom_group(ag) for ag in alt_ags]
assert len(residue_group.atom_groups()) == 1
return hierarchy
def split_conformations(filename, params, log=None):
if log is None: log = Log(verbose=True)
# Read the pdb header - for writing later...
header_contents = get_pdb_header(filename)
# Read in and validate the input file
ens_obj = strip_pdb_to_input(filename, remove_ter=True)
ens_obj.hierarchy.only_model()
# Create a new copy of the structures
new_ens = ens_obj.hierarchy.deep_copy()
# Extract conformers from the structure as set
all_confs = set(ens_obj.hierarchy.altloc_indices())
all_confs.discard('')
if params.options.mode == 'by_residue_name':
sel_resnames = params.options.by_residue_name.resname.split(',')
sel_confs = [
ag.altloc for ag in new_ens.atom_groups()
if (ag.resname in sel_resnames)
]
# List of conformers to output for each structure, and suffixes
out_confs = map(sorted, [
all_confs.intersection(sel_confs),
all_confs.difference(sel_confs)
])
out_suffs = [
params.options.by_residue_name.selected_name,
params.options.by_residue_name.unselected_name
]
elif params.options.mode == 'by_conformer':
sel_resnames = None
sel_confs = None
# One structure for each conformer
out_confs = [[c] for c in sorted(all_confs)]
out_suffs = [''.join(c) for c in out_confs]
elif params.options.mode == 'by_conformer_group':
sel_resnames = None
sel_confs = None
# One structure for each set of supplied conformer sets
out_confs = [
s.split(',') for s in params.options.by_conformer_group.conformers
]
out_suffs = [''.join(c) for c in out_confs]
else:
raise Exception('Invalid selection for options.mode: {}'.format(
params.options.mode))
assert len(out_confs) == len(out_suffs), '{} not same length as {}'.format(
str(out_confs), str(out_suffs))
for confs, suffix in zip(out_confs, out_suffs):
log('Conformers {} -> {}'.format(str(confs), suffix))
# Create paths from the suffixes
out_paths = [
'.'.join([
os.path.splitext(filename)[0], params.output.suffix_prefix, suff,
'pdb'
]) for suff in out_suffs
]
log.subheading('Processing {}'.format(filename[-70:]))
for this_confs, this_path in zip(out_confs, out_paths):
if not this_confs: continue
# Select atoms to keep - no altloc, or altloc in selection
sel_string = ' or '.join(
['altid " "'] + ['altid "{}"'.format(alt) for alt in this_confs])
# Extract selection from the hierarchy
sel_hiery = new_ens.select(
new_ens.atom_selection_cache().selection(sel_string),
copy_atoms=True)
log.bar(True, False)
log('Outputting conformer(s) {} to {}'.format(''.join(this_confs),
this_path))
log.bar()
log('Keeping ANY atom with conformer id: {}'.format(
' or '.join(['" "'] + this_confs)))
log('Selection: \n\t' + sel_string)
if params.options.pruning.prune_duplicates:
log.bar()
log('Pruning redundant conformers')
# Remove an alternate conformers than are duplicated after selection
prune_redundant_alternate_conformations(
hierarchy=sel_hiery,
required_altlocs=[a for a in sel_hiery.altloc_indices() if a],
rmsd_cutoff=params.options.pruning.rmsd_cutoff,
in_place=True,
verbose=params.settings.verbose)
if params.options.reset_altlocs:
log.bar()
# Change the altlocs so that they start from "A"
if len(this_confs) == 1:
conf_hash = {this_confs[0]: ' '}
else:
conf_hash = dict(
zip(this_confs, iotbx.pdb.systematic_chain_ids()))
log('Resetting structure altlocs:')
for k in sorted(conf_hash.keys()):
log('\t{} -> "{}"'.format(k, conf_hash[k]))
if params.settings.verbose: log.bar()
for ag in sel_hiery.atom_groups():
if ag.altloc in this_confs:
if params.settings.verbose:
log('{} -> alt {}'.format(Labeller.format(ag),
conf_hash[ag.altloc]))
ag.altloc = conf_hash[ag.altloc]
if params.options.reset_occupancies:
log.bar()
log('Resetting output occupancies (maximum occupancy of 1.0, etc.)'
)
# Divide through by the smallest occupancy of any complete residues groups with occupancies of less than one
rg_occs = [
calculate_residue_group_occupancy(rg) for rg in
residue_groups_with_complete_set_of_conformers(sel_hiery)
]
non_uni = [v for v in numpy.unique(rg_occs) if 0.0 < v < 1.0]
if non_uni:
div_occ = min(non_uni)
log('Dividing all occupancies by {}'.format(div_occ))
sel_hiery.atoms().set_occ(sel_hiery.atoms().extract_occ() /
div_occ)
# Normalise the occupancies of any residue groups with more than unitary occupancy
log('Fixing any residues that have greater than unitary occupancy')
sanitise_occupancies(hierarchy=sel_hiery,
min_occ=0.0,
max_occ=1.0,
in_place=True,
verbose=params.settings.verbose)
# Perform checks
max_occ = max([
calculate_residue_group_occupancy(rg)
for rg in sel_hiery.residue_groups()
])
log('Maximum occupancy of output structue: {}'.format(max_occ))
assert max_occ >= 0.0, 'maximum occupancy is less than 0.0?!?!'
assert max_occ <= 1.0, 'maximum occupancy is greater than 1.0?!?!'
log.bar()
log('Writing structure: {}'.format(this_path))
log.bar(False, True)
# Write header contents
with open(this_path, 'w') as fh:
fh.write(header_contents)
# Write output file
sel_hiery.write_pdb_file(this_path, open_append=True)
return out_paths
