def create_constraints_file(preminimization_log, outfile_path):
'''This does the work of the convert_to_cst_file.sh script in the Rosetta repository.'''
constraints = []
contents = read_file(preminimization_log)
for line in contents.split('\n'):
if line.startswith("c-alpha"):
line = line.split()
constraints.append("AtomPair CA %s CA %s HARMONIC %s %s" % (line[5], line[7], line[9], line[12]))
write_file(outfile_path, '\n'.join(constraints))
return outfile_path
def create_constraints_file(preminimization_log, outfile_path):
'''This does the work of the convert_to_cst_file.sh script in the Rosetta repository.'''
constraints = []
contents = read_file(preminimization_log)
for line in contents.split('\n'):
if line.startswith("c-alpha"):
line = line.split()
constraints.append("AtomPair CA %s CA %s HARMONIC %s %s" %
(line[5], line[7], line[9], line[12]))
write_file(outfile_path, '\n'.join(constraints))
return outfile_path
def update_pdbs_json():
'''This function was used to update the pdbs.json file to include chain sequences and types.'''
pdb_data = {}
pdb_data_ = json.loads(read_file(os.path.join('..', 'json', 'pdbs.json')))
for k, v in pdb_data_.items():
assert(len(k) == 4)
newk = k.upper()
pdb = PDB(read_file(os.path.join('..', 'pdbs', newk + '.pdb')))
chain_ids = set(pdb.chain_types.keys()).union(set(pdb.seqres_chain_order)).union(set(pdb.atom_sequences.keys()))
v['Chains'] = dict.fromkeys(chain_ids)
for chain_id in chain_ids:
v['Chains'][chain_id] = dict(
Sequence = str(pdb.atom_sequences.get(chain_id)),
Type = pdb.chain_types.get(chain_id),
)
pdb_data[newk] = v
write_file(os.path.join('..', 'json', 'pdbs.json.new'), json.dumps(pdb_data, indent = 4, sort_keys=True))
def update_pdbs_json():
'''This function was used to update the pdbs.json file to include chain sequences and types.'''
pdb_data = {}
pdb_data_ = json.loads(read_file(os.path.join('..', 'json', 'pdbs.json')))
for k, v in pdb_data_.iteritems():
assert(len(k) == 4)
newk = k.upper()
pdb = PDB(read_file(os.path.join('..', 'pdbs', newk + '.pdb')))
chain_ids = set(pdb.chain_types.keys()).union(set(pdb.seqres_chain_order)).union(set(pdb.atom_sequences.keys()))
v['Chains'] = dict.fromkeys(chain_ids)
for chain_id in chain_ids:
v['Chains'][chain_id] = dict(
Sequence = str(pdb.atom_sequences.get(chain_id)),
Type = pdb.chain_types.get(chain_id),
)
pdb_data[newk] = v
write_file(os.path.join('..', 'json', 'pdbs.json.new'), json.dumps(pdb_data, indent = 4, sort_keys=True))
for chain_id, sequence in stripped_pdb.atom_sequences.iteritems():
assert(len(sequence) > 0)
# Check for CSE and MSE
try:
if 'CSE' in stripped_pdb.residue_types:
raise Exception('This case contains a CSE residue which may (or may not) cause an issue with Rosetta depending on the version.')
elif 'MSE' in stripped_pdb.residue_types:
raise Exception('This case contains an MSE residue which may (or may not) cause an issue with Rosetta depending on the version.')
# It looks like MSE (and CSE?) may now be handled - https://www.rosettacommons.org/content/pdb-files-rosetta-format
except Exception, e:
print('%s: %s, chain %s' % (str(e), str(stripped_pdb.pdb_id), chain))
# Turn the lines array back into a valid PDB file
if not(skip_if_exists) or not(os.path.exists(stripped_pdb_path)):
write_file(stripped_pdb_path, '\n'.join(stripped_pdb.lines))
# Create the mapping between PDB and Rosetta residue numbering
# Note: In many Rosetta protocols, '-ignore_unrecognized_res' and '-ignore_zero_occupancy false' are used to allow
# Rosetta to work with structures with missing data and non-canonicals. In those cases, we should supply both flags
# in the string below. Since protocol 16 only uses '-ignore_unrecognized_res', we only use that flag below as otherwise
# we could break the mapping.
rosetta_scripts_bin = os.path.join(settings['local_rosetta_bin'], 'rosetta_scripts%s' % settings['rosetta_binary_type'])
rosetta_database_path = settings['local_rosetta_db_dir']
if not os.path.exists(rosetta_scripts_bin):
raise Exception('The Rosetta scripts executable "{0}" could not be found. Please check your configuration file.'.format(rosetta_database_path))
if not os.path.exists(rosetta_database_path):
raise Exception('The path to the Rosetta database "{0}" could not be found. Please check your configuration file.'.format(rosetta_database_path))
stripped_pdb.construct_pdb_to_rosetta_residue_map(rosetta_scripts_bin,rosetta_database_path, extra_command_flags = '-ignore_unrecognized_res')
atom_to_rosetta_residue_map = stripped_pdb.get_atom_sequence_to_rosetta_json_map()
rosetta_to_atom_residue_map = stripped_pdb.get_rosetta_sequence_to_atom_json_map()
for chain_id, sequence in stripped_pdb.atom_sequences.iteritems():
assert(len(sequence) > 0)
# Check for CSE and MSE
try:
if 'CSE' in stripped_pdb.residue_types:
raise Exception('This case contains a CSE residue which may (or may not) cause an issue with Rosetta depending on the version.')
elif 'MSE' in stripped_pdb.residue_types:
raise Exception('This case contains an MSE residue which may (or may not) cause an issue with Rosetta depending on the version.')
# It looks like MSE (and CSE?) may now be handled - https://www.rosettacommons.org/content/pdb-files-rosetta-format
except Exception, e:
print('%s: %s, chain %s' % (str(e), str(stripped_pdb.pdb_id), chain))
# Turn the lines array back into a valid PDB file
if not(skip_if_exists) or not(os.path.exists(stripped_pdb_path)):
write_file(stripped_pdb_path, '\n'.join(stripped_pdb.lines))
# Create the mapping between PDB and Rosetta residue numbering
# Note: In many Rosetta protocols, '-ignore_unrecognized_res' and '-ignore_zero_occupancy false' are used to allow
# Rosetta to work with structures with missing data and non-canonicals. In those cases, we should supply both flags
# in the string below. Since protocol 16 only uses '-ignore_unrecognized_res', we only use that flag below as otherwise
# we could break the mapping.
rosetta_scripts_bin = os.path.join(settings['local_rosetta_bin'], 'rosetta_scripts%s' % settings['rosetta_binary_type'])
rosetta_database_path = settings['local_rosetta_db_dir']
if not os.path.exists(rosetta_scripts_bin):
raise Exception('The Rosetta scripts executable "{0}" could not be found. Please check your configuration file.'.format(rosetta_database_path))
if not os.path.exists(rosetta_database_path):
raise Exception('The path to the Rosetta database "{0}" could not be found. Please check your configuration file.'.format(rosetta_database_path))
stripped_pdb.construct_pdb_to_rosetta_residue_map(rosetta_scripts_bin,rosetta_database_path, extra_command_flags = '-ignore_unrecognized_res')
atom_to_rosetta_residue_map = stripped_pdb.get_atom_sequence_to_rosetta_json_map()
rosetta_to_atom_residue_map = stripped_pdb.get_rosetta_sequence_to_atom_json_map()
def create_input_files(job_dict,
settings,
pdb_dir_path,
pdb_data_dir,
mutfile_data_dir,
keypair,
dataset_cases,
skip_if_exists=False):
'''Create the stripped PDB files and the mutfiles for the DDG step. Mutfiles are created at this point as we need the
original PDB to generate the residue mapping.
'''
# Read PDB
pdb_id = keypair[0]
chain = keypair[1]
pdb = PDB.from_filepath(pdb_dir_path)
stripped_pdb_path = os.path.join(pdb_data_dir,
'%s_%s.pdb' % (pdb_id, chain))
# Strip the PDB to the list of chains. This also renumbers residues in the PDB for Rosetta.
chains = [chain]
pdb.strip_to_chains(chains)
pdb.strip_HETATMs()
stripped_pdb = PDB('\n'.join(pdb.lines))
# Check to make sure that we haven't stripped all the ATOM lines
if not [line for line in stripped_pdb.lines if line[0:4] == "ATOM"]:
raise Exception("No ATOM lines remain in the stripped PDB file %s." %
stripped_pdb_path)
# Assert that there are no empty sequences
assert (sorted(stripped_pdb.atom_sequences.keys()) == sorted(chains))
for chain_id, sequence in stripped_pdb.atom_sequences.items():
assert (len(sequence) > 0)
# Check for CSE and MSE
try:
if 'CSE' in stripped_pdb.residue_types:
raise Exception(
'This case contains a CSE residue which may (or may not) cause an issue with Rosetta depending on the version.'
)
elif 'MSE' in stripped_pdb.residue_types:
raise Exception(
'This case contains an MSE residue which may (or may not) cause an issue with Rosetta depending on the version.'
)
# It looks like MSE (and CSE?) may now be handled - https://www.rosettacommons.org/content/pdb-files-rosetta-format
except Exception as e:
print(('%s: %s, chain %s' % (str(e), str(stripped_pdb.pdb_id), chain)))
# Turn the lines array back into a valid PDB file
if not (skip_if_exists) or not (os.path.exists(stripped_pdb_path)):
write_file(stripped_pdb_path, '\n'.join(stripped_pdb.lines))
# Create the mapping between PDB and Rosetta residue numbering
# Note: In many Rosetta protocols, '-ignore_unrecognized_res' and '-ignore_zero_occupancy false' are used to allow
# Rosetta to work with structures with missing data and non-canonicals. In those cases, we should supply both flags
# in the string below. Since protocol 16 only uses '-ignore_unrecognized_res', we only use that flag below as otherwise
# we could break the mapping.
rosetta_scripts_bin = os.path.join(
settings['local_rosetta_bin'],
'rosetta_scripts%s' % settings['rosetta_binary_type'])
rosetta_database_path = settings['local_rosetta_db_dir']
if not os.path.exists(rosetta_scripts_bin):
raise Exception(
'The Rosetta scripts executable "{0}" could not be found. Please check your configuration file.'
.format(rosetta_database_path))
if not os.path.exists(rosetta_database_path):
raise Exception(
'The path to the Rosetta database "{0}" could not be found. Please check your configuration file.'
.format(rosetta_database_path))
stripped_pdb.construct_pdb_to_rosetta_residue_map(
rosetta_scripts_bin,
rosetta_database_path,
extra_command_flags='-ignore_unrecognized_res')
atom_to_rosetta_residue_map = stripped_pdb.get_atom_sequence_to_rosetta_json_map(
)
rosetta_to_atom_residue_map = stripped_pdb.get_rosetta_sequence_to_atom_json_map(
)
# Save the PDB <-> Rosetta residue mappings to disk
write_file(
os.path.join(pdb_data_dir,
'%s_%s.rosetta2pdb.resmap.json' % (pdb_id, chain)),
rosetta_to_atom_residue_map)
write_file(
os.path.join(pdb_data_dir,
'%s_%s.pdb2rosetta.resmap.json' % (pdb_id, chain)),
atom_to_rosetta_residue_map)
# Assert that there are no empty sequences in the Rosetta-processed PDB file
total_num_residues = 0
d = json.loads(rosetta_to_atom_residue_map)
for chain_id in chains:
num_chain_residues = len(
[z for z in list(d.values()) if z[0] == chain_id])
total_num_residues += num_chain_residues
assert (num_chain_residues > 0)
# Check that the mutated positions exist and that the wild-type matches the PDB
try:
for dataset_case in dataset_cases:
assert (dataset_case['PDBFileID'] == pdb_id)
# Note: I removed a hack here for 1AJ3->1U5P mapping
# The JSON file does not have the residue IDs in PDB format (5 characters including insertion code) so we need to repad them for the mapping to work
pdb_mutations = [
ChainMutation(mutation['WildTypeAA'],
PDB.ResidueID2String(mutation['ResidueID']),
mutation['MutantAA'],
Chain=mutation['Chain'])
for mutation in dataset_case['Mutations']
]
stripped_pdb.validate_mutations(pdb_mutations)
# Map the PDB mutations to Rosetta numbering which is used by the mutfile format
rosetta_mutations = stripped_pdb.map_pdb_residues_to_rosetta_residues(
pdb_mutations)
if (len(rosetta_mutations) != len(pdb_mutations)) or (None in set(
[m.ResidueID for m in rosetta_mutations])):
raise Exception(
'An error occurred in the residue mapping code for DDG case: %s, %s'
% (pdb_id, pdb_mutations))
# Create the mutfile
mutfile = Mutfile.from_mutagenesis(rosetta_mutations)
mutfilename = os.path.join(
mutfile_data_dir, '%d.mutfile' % (dataset_case['RecordID']))
if os.path.exists(mutfilename):
raise Exception(
'%s already exists. Check that the RecordIDs in the JSON file are all unique.'
% mutfilename)
write_file(
os.path.join(mutfile_data_dir,
'%d.mutfile' % (dataset_case['RecordID'])),
str(mutfile))
except Exception as e:
print((str(e)))
print((traceback.format_exc()))
# Set up --in:file:l parameter
pdb_relpath = os.path.relpath(stripped_pdb_path, settings['output_dir'])
job_dict[os.path.join(
task_subfolder,
'_'.join(keypair))] = dict(input_file_list=[pdb_relpath])
sys.stdout.write('.')
sys.stdout.flush()