def run(self): files = os.listdir(self.path) for f in files: try: with open(os.sep.join([self.path, f]), 'r') as yf: y = yaml.load(yf) ps = PdbStateIdentifier(Structure.objects.get(pdb_code__index=y['pdb'])) ps.run() print(f, y['distance'], y['state'], round(float(ps.activation_value), 2), ps.state) y['distance'] = round(float(ps.activation_value), 2) with open(os.sep.join([self.path, f]), 'w') as syf: yaml.dump(y, syf, indent=4, default_flow_style=False) except Exception as msg: print(f, msg)
def handle(self, *args, **options):
try:
s = Structure.objects.get(pdb_code__index=options['s'])
except:
s = StructureModel.objects.get(protein__entry_name=options['s'], state__slug=options['state'])
if options['gns']:
if s.protein_conformation.protein.family.slug.startswith('002') or s.protein_conformation.protein.family.slug.startswith('003'):
tm2_gn, tm6_gn, tm3_gn, tm7_gn = '2x41', '6x33', '3x44', '7x51'
else:
tm2_gn, tm6_gn, tm3_gn, tm7_gn = '2x41', '6x38', '3x44', '7x52'
for value in options['gns']:
if value.startswith('2'):
tm2_gn = value
elif value.startswith('6'):
tm6_gn = value
elif value.startswith('3'):
tm3_gn = value
elif value.startswith('7'):
tm7_gn = value
if options['cutoffs']:
psi = PdbStateIdentifier(s, tm2_gn=tm2_gn, tm6_gn=tm6_gn, tm3_gn=tm3_gn, tm7_gn=tm7_gn, inactive_cutoff=float(options['cutoffs'][0]), intermediate_cutoff=float(options['cutoffs'][1]), )
else:
psi = PdbStateIdentifier(s, tm2_gn=tm2_gn, tm6_gn=tm6_gn, tm3_gn=tm3_gn, tm7_gn=tm7_gn)
else:
psi = PdbStateIdentifier(s)
psi.run()
print(options['s'], psi.activation_value, psi.state)
def run(self):
strs = Structure.objects.filter(refined=False).exclude(
protein_conformation__protein__parent__family__parent__parent__parent__slug__in
=['001', '004', '005'])
self.match, self.mismatch, self.exceptions = 0, 0, 0
for s in strs:
try:
if self.only_xtals:
psis = PdbStateIdentifier(s, self.tm2_gn, self.tm6_gn,
self.tm3_gn, self.tm7_gn,
self.inact_cutoff,
self.inter_cutoff)
psis.run()
if psis.state != s.state:
print(s, s.state, s.distance, psis.state,
psis.activation_value, 'mismatch')
self.mismatch += 1
else:
print(s, ",", s.state, ',', psis.state,
psis.activation_value, self.inact_cutoff,
self.inter_cutoff)
self.match += 1
else:
r_s = Structure.objects.get(
pdb_code__index=s.pdb_code.index + '_refined')
psi = PdbStateIdentifier(r_s, self.tm2_gn, self.tm6_gn,
self.tm3_gn, self.tm7_gn,
self.inact_cutoff,
self.inter_cutoff)
psi.run()
psis = PdbStateIdentifier(s, self.tm2_gn, self.tm6_gn,
self.tm3_gn, self.tm7_gn,
self.inact_cutoff,
self.inter_cutoff)
psis.run()
if psi.state != psis.state:
print(s, psis.state.slug, psis.activation_value,
psi.state.slug, psi.activation_value)
self.mismatch += 1
else:
self.match += 1
except:
print('Exception: ', s)
self.exceptions += 1
if not self.only_xtals:
hommods = StructureModel.objects.all().exclude(
protein__family__parent__parent__parent__slug__in=[
'001', '004', '005'
])
for h in hommods:
try:
psih = PdbStateIdentifier(h, self.tm2_gn, self.tm6_gn,
self.tm3_gn, self.tm7_gn,
self.inact_cutoff,
self.inter_cutoff)
psih.run()
psiss = PdbStateIdentifier(h.main_template, self.tm2_gn,
self.tm6_gn, self.tm3_gn,
self.tm7_gn, self.inact_cutoff,
self.inter_cutoff)
psiss.run()
if psih.state != psiss.state:
print(h, psiss.state.slug, psiss.activation_value,
psih.state.slug, psih.activation_value)
self.mismatch += 1
else:
self.match += 1
except:
print('Exception hommod:', h)
self.exceptions += 1
print('match:', self.match, 'mismatch:', self.mismatch, 'exceptions:',
self.exceptions)
print('{}-{}-{}-{},{},{},{},{},{}'.format(
self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.mismatch,
self.match, self.exceptions, self.inact_cutoff, self.inter_cutoff))
return 0
def run(self):
strs = Structure.objects.filter(refined=False).exclude(protein_conformation__protein__parent__family__parent__parent__parent__slug__in=['001','004','005'])
self.match, self.mismatch, self.exceptions = 0,0,0
for s in strs:
try:
if self.only_xtals:
psis = PdbStateIdentifier(s, self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.inact_cutoff, self.inter_cutoff)
psis.run()
if psis.state!=s.state:
print(s, s.state, s.distance, psis.state, psis.activation_value, 'mismatch')
self.mismatch+=1
else:
print(s,",",s.state,',', psis.state, psis.activation_value, self.inact_cutoff, self.inter_cutoff)
self.match+=1
else:
r_s = Structure.objects.get(pdb_code__index=s.pdb_code.index+'_refined')
psi = PdbStateIdentifier(r_s, self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.inact_cutoff, self.inter_cutoff)
psi.run()
psis = PdbStateIdentifier(s, self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.inact_cutoff, self.inter_cutoff)
psis.run()
if psi.state!=psis.state:
print(s, psis.state.slug, psis.activation_value, psi.state.slug, psi.activation_value)
self.mismatch+=1
else:
self.match+=1
except:
print('Exception: ', s)
self.exceptions+=1
if not self.only_xtals:
hommods = StructureModel.objects.all().exclude(protein__family__parent__parent__parent__slug__in=['001','004','005'])
for h in hommods:
try:
psih = PdbStateIdentifier(h, self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.inact_cutoff, self.inter_cutoff)
psih.run()
psiss = PdbStateIdentifier(h.main_template, self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.inact_cutoff, self.inter_cutoff)
psiss.run()
if psih.state!=psiss.state:
print(h, psiss.state.slug, psiss.activation_value, psih.state.slug, psih.activation_value)
self.mismatch+=1
else:
self.match+=1
except:
print('Exception hommod:', h)
self.exceptions+=1
print('match:', self.match, 'mismatch:', self.mismatch, 'exceptions:', self.exceptions)
print('{}-{}-{}-{},{},{},{},{},{}'.format(self.tm2_gn, self.tm6_gn, self.tm3_gn, self.tm7_gn, self.mismatch, self.match, self.exceptions, self.inact_cutoff, self.inter_cutoff))
return 0
def new_xtals(self, uniprot):
''' List GPCR crystal structures missing from GPCRdb and the yaml files. Adds missing structures to DB.
'''
structs = self.pdb_request_by_uniprot(uniprot)
try:
protein = Protein.objects.get(accession=uniprot)
except:
protein = None
try:
x50s = Residue.objects.filter(protein_conformation__protein=protein,generic_number__label__in=['1x50','2x50','3x50','4x50','5x50','6x50','7x50'])
except:
x50s = None
if structs!=['null']:
for s in structs:
missing_from_db, missing_yaml = False, False
try:
st_obj = Structure.objects.get(pdb_code__index=s)
except:
if s not in self.exceptions:
check = self.pdb_request_by_pdb(s)
if check==1:
self.db_list.append(s)
missing_from_db = True
if s not in self.yamls and s not in self.exceptions:
if s not in self.db_list:
check = self.pdb_request_by_pdb(s)
else:
check = 1
if check==1:
self.yaml_list.append(s)
missing_yaml = True
if not missing_from_db:
continue
try:
pdb_data_dict = fetch_pdb_info(s, protein, new_xtal=True)
exp_method = pdb_data_dict['experimental_method']
if exp_method=='Electron Microscopy':
st_type = StructureType.objects.get(slug='electron-microscopy')
elif exp_method=='X-ray diffraction':
st_type = StructureType.objects.get(slug='x-ray-diffraction')
if 'deletions' in pdb_data_dict:
for d in pdb_data_dict['deletions']:
presentx50s = []
for x in x50s:
if not d['start']<x.sequence_number<d['end']:
presentx50s.append(x)
# Filter out ones without all 7 x50 positions present in the xtal
if len(presentx50s)!=7:
try:
del self.db_list[self.db_list.index(s)]
missing_from_db = False
del self.yaml_list[self.yaml_list.index(s)]
except:
pass
else:
print('Warning: no deletions in pdb info, check {}'.format(s))
continue
if missing_from_db:
pref_chain = ''
resolution = pdb_data_dict['resolution']
pdb_code, created = WebLink.objects.get_or_create(index=s, web_resource=WebResource.objects.get(slug='pdb'))
pdbl = PDB.PDBList()
pdbl.retrieve_pdb_file(s, pdir='./', file_format="pdb")
with open('./pdb{}.ent'.format(s).lower(),'r') as f:
lines = f.readlines()
pdb_file = ''
publication_date, pubmed, doi = '','',''
state = ProteinState.objects.get(slug='inactive')
new_prot, created = Protein.objects.get_or_create(entry_name=s.lower(), accession=None, name=s.lower(), sequence=pdb_data_dict['wt_seq'], family=protein.family,
parent=protein, residue_numbering_scheme=protein.residue_numbering_scheme,
sequence_type=ProteinSequenceType.objects.get(slug='mod'), source=ProteinSource.objects.get(name='OTHER'),
species=protein.species)
new_prot_conf, created = ProteinConformation.objects.get_or_create(protein=new_prot, state=state, template_structure=None)
for line in lines:
if line.startswith('REVDAT 1'):
publication_date = line[13:22]
if line.startswith('JRNL PMID'):
pubmed = line[19:].strip()
if line.startswith('JRNL DOI'):
doi = line[19:].strip()
pdb_file+=line
pdb_data, created = PdbData.objects.get_or_create(pdb=pdb_file)
d = datetime.strptime(publication_date,'%d-%b-%y')
publication_date = d.strftime('%Y-%m-%d')
try:
if doi!='':
try:
publication = Publication.objects.get(web_link__index=doi)
except Publication.DoesNotExist as e:
p = Publication()
try:
p.web_link = WebLink.objects.get(index=doi, web_resource__slug='doi')
except WebLink.DoesNotExist:
wl = WebLink.objects.create(index=doi,
web_resource = WebResource.objects.get(slug='doi'))
p.web_link = wl
p.update_from_doi(doi=doi)
p.save()
publication = p
elif pubmed!='':
try:
publication = Publication.objects.get(web_link__index=pubmed)
except Publication.DoesNotExist as e:
p = Publication()
try:
p.web_link = WebLink.objects.get(index=pubmed,
web_resource__slug='pubmed')
except WebLink.DoesNotExist:
wl = WebLink.objects.create(index=pubmed,
web_resource = WebResource.objects.get(slug='pubmed'))
p.web_link = wl
p.update_from_pubmed_data(index=pubmed)
p.save()
publication = p
except:
pass
pcs = PdbChainSelector(s, protein)
pcs.run_dssp()
preferred_chain = pcs.select_chain()
# Run state identification
# Create yaml files
with open('../../data/protwis/gpcr/structure_data/constructs/{}.yaml'.format(pdb_code.index), 'w') as construct_file:
yaml.dump({'name': pdb_code.index.lower(), 'protein': protein.entry_name}, construct_file, indent=4)
with open('../../data/protwis/gpcr/structure_data/structures/{}.yaml'.format(pdb_code.index), 'w') as structure_file:
struct_yaml_dict = {'construct': pdb_code.index.lower(), 'pdb': pdb_code.index, 'preferred_chain': preferred_chain, 'auxiliary_protein': '',
'ligand': {'name': 'None', 'pubchemId': 'None', 'title': 'None', 'role': '.nan', 'type': 'None'}, 'signaling_protein': 'None', 'state': 'Inactive'}
auxiliary_proteins, ligands = [], []
if pdb_data_dict['ligands']!='None':
for key, values in pdb_data_dict['ligands'].items():
if key in ['SO4','NA','CLR','OLA','OLB','OLC','TAR','NAG','EPE','BU1','ACM','GOL','PEG','PO4','TLA','BOG','CIT','PLM','BMA','MAN','MLI','PGE']:
continue
else:
ligands.append({'name': key, 'pubchemId': 'None', 'title': pdb_data_dict['ligands'][key]['comp_name'], 'role': '.nan', 'type': 'None'})
for key, values in pdb_data_dict['auxiliary'].items():
if pdb_data_dict['auxiliary'][key]['subtype'] in ['Expression tag', 'Linker']:
continue
else:
auxiliary_proteins.append(pdb_data_dict['auxiliary'][key]['subtype'])
for key, values in pdb_data_dict['construct_sequences'].items():
if key!=protein.entry_name and key not in struct_yaml_dict['auxiliary_protein']:
if 'arrestin' in key:
struct_yaml_dict['signaling_protein'] = key
if len(auxiliary_proteins)>1:
struct_yaml_dict['auxiliary_protein'] = ', '.join(auxiliary_proteins)
if len(ligands)>1:
struct_yaml_dict['ligand'] = ligands
yaml.dump(struct_yaml_dict, structure_file, indent=4, default_flow_style=False)
# Build residue table for structure
build_structure_command = shlex.split('/env/bin/python3 manage.py build_structures -f {}.yaml'.format(pdb_code.index))
subprocess.call(build_structure_command)
# Check state
struct = Structure.objects.get(pdb_code__index=pdb_code.index)
pi = PdbStateIdentifier(struct)
pi.run()
if pi.state!=None:
Structure.objects.filter(pdb_code__index=pdb_code.index).update(state=pi.state)
print(pi.state, pi.activation_value)
with open('../../data/protwis/gpcr/structure_data/structures/{}.yaml'.format(pdb_code.index), 'r') as yf:
struct_yaml = yaml.load(yf)
struct_yaml['state'] = pi.state.name
try:
struct_yaml['distance'] = round(float(pi.activation_value), 2)
except:
struct_yaml['distance'] = None
with open('../../data/protwis/gpcr/structure_data/structures/{}.yaml'.format(pdb_code.index), 'w') as struct_yaml_file:
yaml.dump(struct_yaml, struct_yaml_file, indent=4, default_flow_style=False)
# Check sodium pocket
new_prot_conf.sodium_pocket()
print('{} added to db (preferred_chain chain: {})'.format(s, preferred_chain))
except Exception as msg:
print(s, msg)
def handle(self, *args, **options):
psi = PdbStateIdentifier(options['s'])
psi.run()
print(options['s'], psi.activation_value, psi.state)
def new_xtals(self, uniprot):
''' List GPCR crystal structures missing from GPCRdb and the yaml files. Adds missing structures to DB.
'''
structs = self.pdb_request_by_uniprot(uniprot)
try:
protein = Protein.objects.get(accession=uniprot)
except:
protein = None
try:
x50s = Residue.objects.filter(protein_conformation__protein=protein,generic_number__label__in=['1x50','2x50','3x50','4x50','5x50','6x50','7x50'])
except:
x50s = None
if structs!=['null']:
for s in structs:
missing_from_db, missing_yaml = False, False
try:
st_obj = Structure.objects.get(pdb_code__index=s)
except:
if s not in self.exceptions:
check = self.pdb_request_by_pdb(s)
if check==1:
self.db_list.append(s)
missing_from_db = True
if s not in self.yamls and s not in self.exceptions:
if s not in self.db_list:
check = self.pdb_request_by_pdb(s)
else:
check = 1
if check==1:
self.yaml_list.append(s)
missing_yaml = True
if not missing_from_db:
continue
# try:
pdb_data_dict = fetch_pdb_info(s, protein, new_xtal=True)
exp_method = pdb_data_dict['experimental_method']
if exp_method=='Electron Microscopy':
st_type = StructureType.objects.get(slug='electron-microscopy')
elif exp_method=='X-ray diffraction':
st_type = StructureType.objects.get(slug='x-ray-diffraction')
if 'deletions' in pdb_data_dict:
for d in pdb_data_dict['deletions']:
presentx50s = []
for x in x50s:
if not d['start']<x.sequence_number<d['end']:
presentx50s.append(x)
# Filter out ones without all 7 x50 positions present in the xtal
if len(presentx50s)!=7:
try:
del self.db_list[self.db_list.index(s)]
missing_from_db = False
del self.yaml_list[self.yaml_list.index(s)]
except:
pass
else:
print('Warning: no deletions in pdb info, check {}'.format(s))
continue
if missing_from_db:
pref_chain = ''
resolution = pdb_data_dict['resolution']
pdb_code, created = WebLink.objects.get_or_create(index=s, web_resource=WebResource.objects.get(slug='pdb'))
pdbl = PDB.PDBList()
pdbl.retrieve_pdb_file(s, pdir='./', file_format="pdb")
with open('./pdb{}.ent'.format(s).lower(),'r') as f:
lines = f.readlines()
pdb_file = ''
publication_date, pubmed, doi = '','',''
state = ProteinState.objects.get(slug='inactive')
new_prot, created = Protein.objects.get_or_create(entry_name=s.lower(), accession=None, name=s.lower(), sequence=pdb_data_dict['wt_seq'], family=protein.family,
parent=protein, residue_numbering_scheme=protein.residue_numbering_scheme,
sequence_type=ProteinSequenceType.objects.get(slug='mod'), source=ProteinSource.objects.get(name='OTHER'),
species=protein.species)
new_prot_conf, created = ProteinConformation.objects.get_or_create(protein=new_prot, state=state, template_structure=None)
for line in lines:
if line.startswith('REVDAT 1'):
publication_date = line[13:22]
if line.startswith('JRNL PMID'):
pubmed = line[19:].strip()
if line.startswith('JRNL DOI'):
doi = line[19:].strip()
pdb_file+=line
pdb_data, created = PdbData.objects.get_or_create(pdb=pdb_file)
d = datetime.strptime(publication_date,'%d-%b-%y')
publication_date = d.strftime('%Y-%m-%d')
try:
if doi!='':
try:
publication = Publication.objects.get(web_link__index=doi)
except Publication.DoesNotExist as e:
p = Publication()
try:
p.web_link = WebLink.objects.get(index=doi, web_resource__slug='doi')
except WebLink.DoesNotExist:
wl = WebLink.objects.create(index=doi,
web_resource = WebResource.objects.get(slug='doi'))
p.web_link = wl
p.update_from_doi(doi=doi)
p.save()
publication = p
elif pubmed!='':
try:
publication = Publication.objects.get(web_link__index=pubmed)
except Publication.DoesNotExist as e:
p = Publication()
try:
p.web_link = WebLink.objects.get(index=pubmed,
web_resource__slug='pubmed')
except WebLink.DoesNotExist:
wl = WebLink.objects.create(index=pubmed,
web_resource = WebResource.objects.get(slug='pubmed'))
p.web_link = wl
p.update_from_pubmed_data(index=pubmed)
p.save()
publication = p
except:
pass
pcs = PdbChainSelector(s, protein)
pcs.run_dssp()
preferred_chain = pcs.select_chain()
# Run state identification
# Create yaml files
with open(os.sep.join([settings.DATA_DIR, 'structure_data','constructs', '{}.yaml'.format(pdb_code.index)]), 'w') as construct_file:
yaml.dump({'name': pdb_code.index.lower(), 'protein': protein.entry_name}, construct_file, indent=4)
with open(os.sep.join([settings.DATA_DIR, 'structure_data','structures','{}.yaml'.format(pdb_code.index)]), 'w') as structure_file:
struct_yaml_dict = {'construct': pdb_code.index.lower(), 'pdb': pdb_code.index, 'preferred_chain': preferred_chain, 'auxiliary_protein': '',
'ligand': {'name': 'None', 'pubchemId': 'None', 'title': 'None', 'role': '.nan', 'type': 'None'}, 'signaling_protein': 'None', 'state': 'Inactive'}
auxiliary_proteins, ligands = [], []
if pdb_data_dict['ligands']!='None':
for key, values in pdb_data_dict['ligands'].items():
if key in ['SO4','NA','CLR','OLA','OLB','OLC','TAR','NAG','EPE','BU1','ACM','GOL','PEG','PO4','TLA','BOG','CIT','PLM','BMA','MAN','MLI','PGE','SIN','PGO','MES','ZN','NO3','NI','MG','PG4']:
continue
else:
ligands.append({'name': key, 'pubchemId': 'None', 'title': pdb_data_dict['ligands'][key]['comp_name'], 'role': '.nan', 'type': 'None'})
for key, values in pdb_data_dict['auxiliary'].items():
if pdb_data_dict['auxiliary'][key]['subtype'] in ['Expression tag', 'Linker']:
continue
else:
auxiliary_proteins.append(pdb_data_dict['auxiliary'][key]['subtype'])
for key, values in pdb_data_dict['construct_sequences'].items():
if key!=protein.entry_name and key not in struct_yaml_dict['auxiliary_protein']:
if 'arrestin' in key:
struct_yaml_dict['signaling_protein'] = key
if len(auxiliary_proteins)>1:
struct_yaml_dict['auxiliary_protein'] = ', '.join(auxiliary_proteins)
if len(ligands)>1:
struct_yaml_dict['ligand'] = ligands
yaml.dump(struct_yaml_dict, structure_file, indent=4, default_flow_style=False)
# Build residue table for structure
build_structure_command = shlex.split('/env/bin/python3 manage.py build_structures -f {}.yaml'.format(pdb_code.index))
subprocess.call(build_structure_command)
# Check state
struct = Structure.objects.get(pdb_code__index=pdb_code.index)
pi = PdbStateIdentifier(struct)
pi.run()
if pi.state!=None:
Structure.objects.filter(pdb_code__index=pdb_code.index).update(state=pi.state)
print(pi.state, pi.activation_value)
with open('../../data/protwis/gpcr/structure_data/structures/{}.yaml'.format(pdb_code.index), 'r') as yf:
struct_yaml = yaml.load(yf)
struct_yaml['state'] = pi.state.name
try:
struct_yaml['distance'] = round(float(pi.activation_value), 2)
except:
struct_yaml['distance'] = None
with open('../../data/protwis/gpcr/structure_data/structures/{}.yaml'.format(pdb_code.index), 'w') as struct_yaml_file:
yaml.dump(struct_yaml, struct_yaml_file, indent=4, default_flow_style=False)
# Check sodium pocket
new_prot_conf.sodium_pocket()
print('{} added to db (preferred_chain chain: {})'.format(s, preferred_chain))