You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
88 lines
2.8 KiB
Python
88 lines
2.8 KiB
Python
import requests
|
|
import json
|
|
from drugtools.env_setup import ENV,postgres_conn
|
|
from psycopg2 import extras
|
|
from collections import namedtuple
|
|
from tqdm import tqdm
|
|
|
|
|
|
RecordStuff = namedtuple("RecordStuff", "nct_id condition ui uri rootSource name")
|
|
|
|
class Requestor():
|
|
def __init__(self,api_key):
|
|
self.key = api_key
|
|
|
|
def search(self,search_term,inputType="sourceUi", returnIdType="code", addnl_terms={}):
|
|
query_terms = {
|
|
"apiKey":self.key,
|
|
"sabs":"ICD10",
|
|
"string":search_term,
|
|
"returnIdType":returnIdType,
|
|
"inputType":inputType
|
|
} | addnl_terms
|
|
query = "https://uts-ws.nlm.nih.gov/rest/search/current/"
|
|
|
|
r = requests.get(query,params=query_terms)
|
|
return r
|
|
|
|
|
|
r = Requestor(ENV.get("UMLS_API_KEY"))
|
|
|
|
|
|
with postgres_conn() as pconn, pconn.cursor(cursor_factory=extras.DictCursor) as pcurse:
|
|
sql = """
|
|
select nct_id, downcase_mesh_term
|
|
from ctgov.browse_conditions bc
|
|
where
|
|
mesh_type = 'mesh-list'
|
|
and
|
|
nct_id in (select distinct nct_id from history.trial_snapshots ts)
|
|
order by nct_id
|
|
;
|
|
"""
|
|
sql2 = """
|
|
with cte as (
|
|
/* Keywords added too much noise
|
|
select nct_id,downcase_name
|
|
from ctgov.keywords k
|
|
where nct_id in (select distinct nct_id from history.trial_snapshots ts)
|
|
union */
|
|
select nct_id, downcase_name
|
|
from ctgov.conditions c
|
|
union
|
|
select nct_id ,downcase_mesh_term as downcase_name
|
|
from ctgov.browse_conditions bc
|
|
where mesh_type = 'mesh-list'
|
|
)
|
|
select nct_id, downcase_name from cte
|
|
where nct_id in (select distinct nct_id from history.trial_snapshots ts)
|
|
order by nct_id
|
|
"""
|
|
pcurse.execute(sql2)
|
|
rows = pcurse.fetchall()
|
|
|
|
entries = []
|
|
|
|
for row in tqdm(rows,desc="Search MeSH terms"):
|
|
nctid = row[0]
|
|
condition = row[1]
|
|
# print(nctid,condition)
|
|
|
|
results = r.search(row[1]).json().get('result', Exception("No result entry in json")).get('results',Exception("No results entry in json"))
|
|
|
|
#if results are empty?
|
|
if not results:
|
|
entries.append(RecordStuff(nctid,condition,None,None,None,None))
|
|
else:
|
|
for entry in results:
|
|
entries.append(RecordStuff(nctid, condition, entry["ui"], entry["uri"], entry["rootSource"], entry["name"]))
|
|
|
|
sql_insert = """
|
|
INSERT INTO "DiseaseBurden".trial_to_icd10
|
|
(nct_id, "condition", ui,uri,rootsource,"name","source",approved,approval_timestamp)
|
|
VALUES
|
|
(%(nct_id)s, %(condition)s, %(ui)s, %(uri)s, %(rootSource)s, %(name)s, 'UMLS API search', null,null)
|
|
"""
|
|
for entry in tqdm(entries,desc="Inserting entries to DB"):
|
|
pcurse.execute(sql_insert,entry._asdict())
|