def __init__(self, xmlQuery):
results = post_query(xmlQuery)
self.entityLevel = (len(results) > 0) and (":" in results[0])
self.structureIds = list(set(results))
self.exclusive = False
def __init__(self,
smiles,
match_type=SUBSTRUCTURE_STEREOSPECIFIC,
percentSimilarity=0.0):
'''Constructor to setup filter that matches any entry with at least one
chemical component that matches the specified SMILES string using the
specified query type.
For details see:
`Chemical Structure Search <http://www.rcsb.org/pdb/staticHelp.do?p=help/advancedsearch/chemSmiles.html>`_
Parameters
----------
smiles : str
SMILES string representing chemical structure
match_type : str
One of the 5 supported types
percentSimilarity : float
percent similarity for similarity search. This parameter is ignored
for all other query types [default: 0.0]
'''
max_rows = 1000
query = ('{'
'"query": {'
'"type": "terminal",'
'"service": "chemical",'
'"parameters": {'
f'"value": "{smiles}",'
'"type": "descriptor",'
'"descriptor_type": "SMILES",'
f'"match_type": "{match_type}"'
'}'
'},'
'"return_type": "entry",'
'"request_options": {'
'"pager": {'
'"start": 0,'
f'"rows": {max_rows}'
'},'
'"scoring_strategy": "combined",'
'"sort": ['
'{'
'"sort_by": "score",'
'"direction": "desc"'
'}'
']'
'}'
'}')
result_type, identifiers, scores = post_query(query)
self.result_type = result_type
self.structureIds = set()
for identifier, score in zip(identifiers, scores):
if (score * 100.0 >= percentSimilarity):
self.structureIds.add(identifier)
def __init__(self, query):
result_type, results, scores = post_query(query)
self.result_type = result_type
#self.entityLevel = (len(results) > 0) and (":" in results[0])
self.entityLevel = result_type == 'polymer_entity'
#print('result_type:', result_type, 'entityLevel:', self.entityLevel)
self.structureIds = list(set(results))
#print('structureIds:', self.structureIds)
self.exclusive = False
def get_dataset(xmlQuery):
"""
Runs an RCSB PDB Advanced Search web service using an XML query description.
See https://www.rcsb.org/pdb/staticHelp.do?p=help/advancedSearch.html Advanced Search
The returned dataset contains the following field dependent on the query type:
# structureId, e.g., 1STP
# structureChainId, e.g., 4HHB.A
# ligandId, e.g., HEM
:param xmlQuery: RCSB PDB advanced query xml string
:return: dataset with matching ids
"""
# run advanced query
ids = post_query(xmlQuery)
# convert list of ids to a list of lists (required for dataframe creation below)
id_list = [[i] for i in ids]
# convert list of lists to a dataframe
spark = SparkSession.builder.getOrCreate()
# distinguish 3 types of results based on length of string
# structureId: 4 (e.g., 4HHB)
# structureEntityId: > 4 (e.g., 4HHB:1)
# entityId: < 4 (e.g., HEM)
if len(ids[0]) > 4:
ds: DataFrame = spark.createDataFrame(id_list, ['pdbEntityId'])
# if results contain an entity id, e.g., 101M:1, then map entityId to pdbChainId
ds = ds.withColumn("pdbId", substring_index(ds.pdbEntityId, ':', 1))
ds = ds.withColumn("entityId", substring_index(ds.pdbEntityId, ':',
-1))
mapping = __get_entity_to_chain_id()
ds = ds.join(mapping, (ds.pdbId == mapping.structureId) &
(ds.entityId == mapping.entity_id))
ds = ds.select(ds.pdbChainId)
elif len(ids[0]) < 4:
ds: DataFrame = spark.createDataFrame(id_list, ['ligandId'])
else:
ds: DataFrame = spark.createDataFrame(id_list, ['pdbId'])
return ds