import requests
import logging
from ..utils.errors import *
logger = logging.getLogger(__name__)
def search_pubchem(keyword):
keyword = keyword.replace(';', '%3B')
url = 'https://pubchem.ncbi.nlm.nih.gov/sdq/sdqagent.cgi?infmt=json&outfmt=json&query={%22select%22:%22*%22,%22collection%22:%22compound%22,%22order%22:[%22relevancescore,desc%22],%22start%22:1,%22limit%22:10,%22where%22:{%22ands%22:[{%22*%22:%22' + keyword + '%22}]},%22width%22:1000000,%22listids%22:0}'
data = requests.get(url).json()
return data
def pubchem_iupac2cid(iupac, strict=False):
iupac = iupac.replace('λ', 'lambda')
data = search_pubchem(iupac)
try:
rows = data['SDQOutputSet'][0]['rows']
except KeyError as e:
raise ChemMCPSearchFailError("Cannot find a molecule/compound on PubChem that matches the input IUPAC name.") from e
potential_cid = None
if len(rows) > 0: # If there are rows, meaning that there are some results
potential_cid = rows[0]['cid'] # Use the first one as the last resort
# Check if the IUPAC name matches anyone in the result
matched_item = None
iupac_lower = iupac.lower()
for item in rows:
try:
if item['iupacname'].lower() == iupac_lower:
matched_item = item
break
except KeyError:
continue
if matched_item is not None:
return matched_item['cid']
# if len(rows) == 0 or matched_item is None, check if each component of the IUPAC name
parts = iupac.split(';')
if len(parts) > 1:
logger.info('The input IUPAC name contains multiple molecules/compounds. Searching for each of them.')
parts_cid = []
parts_cannot_find = []
for part in parts:
part = part.strip()
try:
cid = pubchem_iupac2cid(part)
except ChemMCPSearchFailError:
parts_cannot_find.append(part)
else:
parts_cid.append(cid)
if len(parts_cannot_find) > 0:
raise ChemMCPSearchFailError("Cannot find a molecule/compound for the following parts of the input IUPAC name: %s" % ', '.join(parts_cannot_find))
else:
return tuple(parts_cid)
else:
if potential_cid is not None:
if strict:
raise ChemMCPSearchFailError("Cannot find a molecule/compound that matches the input IUPAC name.")
logger.info("Cannot find a molecule/compound that matches the input IUPAC name. Using the first matched one.")
return potential_cid
raise ChemMCPSearchFailError("Cannot find a molecule/compound on PubChem that matches the input IUPAC name. Possible reasons: 1) The input must be a valid IUPAC name. 2) Must input only one at a time. If you have multiple molecules to look up, please input each of them at a time.")
def pubchem_name2cid_old(name):
data = search_pubchem(name)
try:
rows = data['SDQOutputSet'][0]['rows']
except KeyError as e:
raise ChemMCPSearchFailError("Cannot find a molecule/compound that matches the input IUPAC name.") from e
if len(rows) > 0:
if len(rows) > 1:
logger.info("There are more than one molecules/compounds that match the input name. Using the first matched one.")
return rows[0]['cid']
else:
raise ChemMCPSearchFailError("Cannot find a molecule/compound that matches the input name.")
def pubchem_name2cid(name):
url = 'https://pubchem.ncbi.nlm.nih.gov/rest/pug/compound/name/' + name + '/cids/JSON'
data = requests.get(url).json()
try:
cid = data['IdentifierList']['CID'][0]
except KeyError as e:
raise ChemMCPSearchFailError("Cannot find a molecule/compound that matches the input chemical name.") from e
return cid
