def remap_go_ids(stmts):
for stmt in stmts:
for agent in stmt.agent_list():
if agent is not None and 'GO' in agent.db_refs:
prim_id = go_client.get_primary_id(agent.db_refs['GO'])
if prim_id:
agent.db_refs['GO'] = prim_id
def _get_db_refs(entity_term):
db_refs = {}
for xr in entity_term['xrefs']:
ns = xr['namespace']
if ns == 'uniprot':
db_refs['UP'] = xr['id']
elif ns == 'hgnc':
db_refs['HGNC'] = xr['id']
elif ns == 'pfam':
fplx_id = famplex_map.get(('PF', xr['id']))
if fplx_id:
db_refs['FPLX'] = fplx_id
db_refs['PF'] = xr['id']
elif ns == 'interpro':
fplx_id = famplex_map.get(('IP', xr['id']))
if fplx_id:
db_refs['FPLX'] = fplx_id
db_refs['IP'] = xr['id']
elif ns == 'chebi':
db_refs['CHEBI'] = xr['id']
elif ns == 'pubchem':
db_refs['PUBCHEM'] = xr['id']
elif ns == 'go':
go_id = xr['id']
# Handle secondary to primary mapping if necessary
pri = go_client.get_primary_id(go_id)
if pri:
go_id = pri
db_refs['GO'] = go_id
elif ns == 'mesh':
db_refs['MESH'] = xr['id']
elif ns == 'hmdb':
db_refs['HMDB'] = xr['id']
elif ns == 'simple_chemical':
if xr['id'].startswith('HMDB'):
db_refs['HMDB'] = xr['id']
# We handle "be" here for compatibility with older versions
elif ns in ('fplx', 'be'):
db_refs['FPLX'] = xr['id']
# These name spaces are ignored
elif ns in ['uaz']:
pass
else:
logger.warning('Unhandled xref namespace: %s' % ns)
db_refs['TEXT'] = entity_term['text']
db_refs = standardize_db_refs(db_refs)
return db_refs
def test_go_secondary_to_primary():
assert go_client.get_primary_id('GO:0007067') == 'GO:0000278'
def _get_db_refs(entity_term, organism_priority=None):
db_refs = {}
for xr in entity_term['xrefs']:
ns = xr['namespace']
if ns == 'uniprot':
# Note: we add both full protein and protein chain
# IDs here so that we can appli organism prioritization in
# a uniform way. Later these will be separated out.
up_id = xr['id']
db_refs['UP'] = up_id
elif ns == 'hgnc':
db_refs['HGNC'] = xr['id']
elif ns == 'pfam':
fplx_id = famplex_map.get(('PF', xr['id']))
if fplx_id:
db_refs['FPLX'] = fplx_id
db_refs['PF'] = xr['id']
elif ns == 'interpro':
fplx_id = famplex_map.get(('IP', xr['id']))
if fplx_id:
db_refs['FPLX'] = fplx_id
db_refs['IP'] = xr['id']
elif ns == 'chebi':
db_refs['CHEBI'] = xr['id']
elif ns == 'pubchem':
db_refs['PUBCHEM'] = xr['id']
elif ns == 'go':
go_id = xr['id']
# Handle secondary to primary mapping if necessary
pri = go_client.get_primary_id(go_id)
if pri:
go_id = pri
db_refs['GO'] = go_id
elif ns == 'mesh':
db_refs['MESH'] = xr['id']
elif ns == 'hmdb':
db_refs['HMDB'] = xr['id']
elif ns == 'simple_chemical':
if xr['id'].startswith('HMDB'):
db_refs['HMDB'] = xr['id']
# We handle "be" here for compatibility with older versions
elif ns in ('fplx', 'be'):
db_refs['FPLX'] = xr['id']
elif ns == 'proonto':
db_refs['PR'] = xr['id']
# These name spaces are ignored
elif ns in ['uaz']:
pass
else:
logger.warning('Unhandled xref namespace: %s' % ns)
db_refs['TEXT'] = entity_term['text']
# If we have a UniProt grounding and we have a non-default
# organism priority list, we call the prioritization function
if db_refs.get('UP'):
if organism_priority:
# These are all the unique groundings in the alt-xrefs list,
# which redundantly lists the same match multiple times because
# it enumerates multiple synonyms for organisms redundantly
unique_altxrefs = \
set((axr['namespace'], axr['id'])
for axr in entity_term.get('alt-xrefs', []))
# This returns a single prioritized UniProt ID or None
prioritized_id = \
prioritize_organism_grounding(db_refs['UP'],
unique_altxrefs,
organism_priority)
# If we got an ID, we set the UP grounding to that, otherwise
# we keep what we already got from the primary xref
if prioritized_id:
db_refs['UP'] = prioritized_id
# After all this, we need to separate protein chain grounding
# and so if we are dealing with one of those, we pop out the UP
# key, split the ID to get the chain ID and add that in the UPPRO
# namespace.
if '#' in db_refs['UP']:
up_id = db_refs.pop('UP', None)
db_refs['UPPRO'] = up_id.split('#')[1]
db_refs = standardize_db_refs(db_refs)
return db_refs
def _get_db_refs(entity_term):
agent_name = entity_term['text']
db_refs = {}
for xr in entity_term['xrefs']:
ns = xr['namespace']
if ns == 'uniprot':
up_id = xr['id']
db_refs['UP'] = up_id
# Look up official names in UniProt
gene_name = up_client.get_gene_name(up_id)
if gene_name is not None:
agent_name = gene_name
# If the gene name corresponds to an HGNC ID, add it to the
# db_refs
if up_client.is_human(up_id):
hgnc_id = hgnc_client.get_hgnc_id(gene_name)
if hgnc_id:
db_refs['HGNC'] = hgnc_id
elif ns == 'hgnc':
hgnc_id = xr['id']
db_refs['HGNC'] = hgnc_id
# Look up the standard gene symbol and set as name
hgnc_name = hgnc_client.get_hgnc_name(hgnc_id)
if hgnc_name:
agent_name = hgnc_name
# Look up the corresponding uniprot id
up_id = hgnc_client.get_uniprot_id(hgnc_id)
if up_id:
db_refs['UP'] = up_id
elif ns == 'pfam':
be_id = famplex_map.get(('PF', xr['id']))
if be_id:
db_refs['FPLX'] = be_id
agent_name = be_id
db_refs['PF'] = xr['id']
elif ns == 'interpro':
be_id = famplex_map.get(('IP', xr['id']))
if be_id:
db_refs['FPLX'] = be_id
agent_name = be_id
db_refs['IP'] = xr['id']
elif ns == 'chebi':
db_refs['CHEBI'] = xr['id']
elif ns == 'pubchem':
db_refs['PUBCHEM'] = xr['id']
elif ns == 'go':
go_id = xr['id']
# Handle secondary to primary mapping if necessary
pri = go_client.get_primary_id(go_id)
if pri:
go_id = pri
db_refs['GO'] = go_id
elif ns == 'mesh':
db_refs['MESH'] = xr['id']
elif ns == 'hmdb':
db_refs['HMDB'] = xr['id']
elif ns == 'simple_chemical':
if xr['id'].startswith('HMDB'):
db_refs['HMDB'] = xr['id']
elif ns == 'be':
db_refs['FPLX'] = xr['id']
agent_name = db_refs['FPLX']
# These name spaces are ignored
elif ns in ['uaz']:
pass
else:
logger.warning('Unhandled xref namespace: %s' % ns)
db_refs['TEXT'] = entity_term['text']
return agent_name, db_refs