def get_metadata(job_id):
"""
Get the metadata of the model, created by the modeling job.
:param jobid: the job_id returned by 'submit'
:return: The json metadata object. If the job status is not SUCCESS,
this method returns an error.
"""
from hommod.application import celery
result = celery.AsyncResult(job_id)
if result.status != 'SUCCESS':
return jsonify({'error': "{} has status {}".format(job_id, result.status)}), 400
try:
path = result.get()
except Exception:
return jsonify({'error': result.traceback}), 500
if path is None:
message = 'Job %s finished, but without creating a model. This could be due to lack of a suitable template.' % job_id
return jsonify({'error': message}), 500
try:
data = {}
data['selected_targets'] = model_storage.extract_selected_targets(path)
data['alignments'] = [alignment.as_dict()
for alignment in model_storage.extract_alignments(path)]
return jsonify(data)
except:
return jsonify({'error': traceback.format_exc()}), 500
def test_create_model_cox():
from hommod.tasks import create_model
path = create_model(
"MLATRVFSLVGKRAISTSVCVRAHESVVKSEDFSLPAYMDRRDHPLPEVAHVKHLSASQKALKEKEKASWSS" +
"LSMDEKVELYRIKFKESFAEMNRGSNEWKTVVGGAMFFIGFTALVIMWQKHYVYGPLPQSFDKEWVAKQTKR" +
"MLDMKVNPIQGLASKWDYEKNEWKK", "HUMAN", None, TemplateID('2y69', 'D'))
ok_(path is not None)
name = os.path.splitext(os.path.basename(path))[0]
pdb_name = os.path.join(name, 'target.pdb')
with tarfile.open(path) as tf:
ok_(pdb_name in tf.getnames())
alignments = model_storage.extract_alignments(path)
ok_(len(alignments) >= 13)
def select_best_model(tar_paths):
best_path = None
best_pid = 0.0
for tar_path in tar_paths:
model_name = model_storage.get_model_name_from_path(tar_path)
sequence_id = model_storage.get_sequence_id_from_name(model_name)
pdb_str = model_storage.extract_model(tar_path)
chain_order = parse_chain_order_from_string(pdb_str)
chain_alignments = model_storage.extract_alignments(tar_path)
if len(chain_order) != len(chain_alignments):
_log.error("{}: {} chains in pdb file, vs. {} alignments in fasta"
.format(tar_path, len(chain_order), len(chain_alignments)))
continue
chain_targets = model_storage.extract_selected_targets(tar_path)
chain_ids = list(filter(lambda chain_id: chain_targets[chain_id] == sequence_id,
chain_targets.keys()))
if len(chain_ids) <= 0:
_log.error("no main target {} in selected targets of {}".format(sequence_id, tar_path))
continue
if chain_ids[0] not in chain_order:
_log.error("chain {} not in {}".format(chain_ids[0], tar_path))
continue
chain_i = chain_order.index(chain_ids[0])
alignment = chain_alignments[chain_i]
keys = list(alignment.aligned_sequences.keys())
if len(keys) != 2:
_log.error("alignment for chain {} in {} has too many keys: {}".format(chain_ids[0], tar_path, keys))
continue
pid = alignment.get_percentage_identity(keys[0], keys[1])
if pid > best_pid:
best_pid = pid
best_path = tar_path
return best_path
def get_metadata_by_model_id(model_id):
"""
Get the metadata of the model, created by the modeling job.
:param model_id: the id returned by 'get_model_if_exists'
:return: The json metadata object.
"""
path = model_storage.get_tar_path_from_name(model_id)
if not os.path.isfile(path):
return jsonify({'error': "no such model"}), 400
try:
data = {}
data['selected_targets'] = model_storage.extract_selected_targets(path)
data['alignments'] = [alignment.as_dict()
for alignment in model_storage.extract_alignments(path)]
return jsonify(data)
except:
return jsonify({'error': traceback.format_exc()}), 500
def get_metadata_by_model_id(model_id):
"""
Get the metadata of the model, created by the modeling job.
:param model_id: the id returned by 'get_model_if_exists'
:return: The json metadata object.
"""
path = model_storage.get_tar_path_from_name(model_id)
if not os.path.isfile(path):
return jsonify({'error': "no such model"}), 400
try:
data = {}
data['selected_targets'] = model_storage.extract_selected_targets(path)
data['alignments'] = [
alignment.as_dict()
for alignment in model_storage.extract_alignments(path)
]
return jsonify(data)
except:
return jsonify({'error': traceback.format_exc()}), 500
def select_best_model(tar_paths):
best_path = None
best_pid = 0.0
for tar_path in tar_paths:
model_name = model_storage.get_model_name_from_path(tar_path)
sequence_id = model_storage.get_sequence_id_from_name(model_name)
pdb_str = model_storage.extract_model(tar_path)
chain_order = parse_chain_order_from_string(pdb_str)
chain_alignments = model_storage.extract_alignments(tar_path)
chain_targets = model_storage.extract_selected_targets(tar_path)
chain_ids = list(filter(lambda chain_id: chain_targets[chain_id] == sequence_id,
chain_targets.keys()))
if len(chain_ids) <= 0:
_log.error("no main target {} in selected targets of {}".format(sequence_id, tar_path))
continue
chain_i = chain_order.index(chain_ids[0])
if chain_i == -1:
_log.error("chain {} not in {}".format(chain_ids[0], tar_path))
continue
alignment = chain_alignments[chain_i]
keys = alignment.aligned_sequences.keys()
if len(keys) != 2:
_log.error("alignment for chain {} in {} has too many keys: {}".format(chain_ids[0], tar_path, keys))
continue
pid = alignment.get_percentage_identity(keys[0], keys[1])
if pid > best_pid:
best_pid = pid
best_path = tar_path
return best_path
def get_metadata(job_id):
"""
Get the metadata of the model, created by the modeling job.
:param jobid: the job_id returned by 'submit'
:return: The json metadata object. If the job status is not SUCCESS,
this method returns an error.
"""
from hommod.application import celery
result = celery.AsyncResult(job_id)
if result.status != 'SUCCESS':
return jsonify(
{'error': "{} has status {}".format(job_id, result.status)}), 400
try:
path = result.get()
except Exception:
return jsonify({'error': result.traceback}), 500
if path is None:
message = 'Job %s finished, but without creating a model. This could be due to lack of a suitable template.' % job_id
return jsonify({'error': message}), 500
try:
data = {}
data['selected_targets'] = model_storage.extract_selected_targets(path)
data['alignments'] = [
alignment.as_dict()
for alignment in model_storage.extract_alignments(path)
]
return jsonify(data)
except:
return jsonify({'error': traceback.format_exc()}), 500