def associated_events(self, drugname, sortby="COUNT"):
""" This method will return a sorted list of drug-event frequencies.
"""
drugname = drugs.standardize([drugname])[0]
sql = "SELECT DRUGNAME, PT, COUNT FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = '%s' ORDER BY %s DESC" % (drugname, sortby)
results = self.conn.execute(sql).fetchall()
return [[str(k),str(v),c] for k,v,c in results]
def associated_events(self, drugname, sortby="COUNT"):
""" This method will return a sorted list of drug-event frequencies.
"""
drugname = drugs.standardize([drugname])[0]
sql = "SELECT DRUGNAME, PT, COUNT FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = '%s' ORDER BY %s DESC" % (
drugname, sortby)
results = self.conn.execute(sql).fetchall()
return [[str(k), str(v), c] for k, v, c in results]
def mine_prr_by_drug(self, drug, n=3):
""" Given a drug, compute PRR for all events assocated with drug for
drug-event with frequency >= n.
"""
drug = drugs.standardize([drug])[0]
sql = "SELECT PT FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = '%s' AND COUNT >= %s" % (drug, str(n))
result = self.conn.execute(sql).fetchall()
events = [str(i[0]) for i in result]
prr_list = [self.prr(drug, e) for e in events]
prr_list = [i for i in prr_list if i['PRR'] != None]
idx = sorted(range(len(prr_list)), key = lambda k: -prr_list[k]['PRR'])
prr_list = [prr_list[i] for i in idx]
return prr_list
def mine_prr_by_drug(self, drug, n=3):
""" Given a drug, compute PRR for all events assocated with drug for
drug-event with frequency >= n.
"""
drug = drugs.standardize([drug])[0]
sql = "SELECT PT FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = '%s' AND COUNT >= %s" % (
drug, str(n))
result = self.conn.execute(sql).fetchall()
events = [str(i[0]) for i in result]
prr_list = [self.prr(drug, e) for e in events]
prr_list = [i for i in prr_list if i['PRR'] != None]
idx = sorted(range(len(prr_list)), key=lambda k: -prr_list[k]['PRR'])
prr_list = [prr_list[i] for i in idx]
return prr_list
def drug_event_stats(self, drug, event):
""" This method computes frequencies used in calculating the PRR:
----------------------------------+--------------------------------------------+
| Drug of interest | All other drugs | Total |
----------------------------------+------------------+-----------------+-------|
Adverse drug reaction of interest | A | B | A + B |
----------------------------------+------------------+-----------------+-------|
All other adverse drug reactions | C | D | C + D |
----------------------------------+------------------+-----------------+-------|
Total | A+C | B+D |A+B+C+D|
-------------------------------------------------------------------------------+
"""
# print """
#+-----------------------------------+------------------+-------------------------+
#| | Drug of interest | All other drugs | Total |
#+-----------------------------------+------------------+-----------------+-------|
#| Adverse drug reaction of interest | A | B | A + B |
#+-----------------------------------+------------------+-----------------+-------|
#| All other adverse drug reactions | C | D | C + D |
#+-----------------------------------+------------------+-----------------+-------|
#| Total | A+C | B+D |A+B+C+D|
#+-----------------------------------+------------------+-----------------+-------+
#"""
drug = drugs.standardize([drug])[0]
event = event.upper()
A = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = \"%s\" AND PT = \"%s\"" % (
drug, event)
B = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME <> \"%s\" AND PT = \"%s\"" % (
drug, event)
C = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = \"%s\" AND PT <> \"%s\"" % (
drug, event)
D = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME <> \"%s\" AND PT <> \"%s\"" % (
drug, event)
A = self.conn.execute(A).fetchone()[0]
B = self.conn.execute(B).fetchone()[0]
C = self.conn.execute(C).fetchone()[0]
D = self.conn.execute(D).fetchone()[0]
return {"A": A, "B": B, "C": C, "D": D, "drug": drug, "event": event}
def drug_event_stats(self, drug, event):
""" This method computes frequencies used in calculating the PRR:
----------------------------------+--------------------------------------------+
| Drug of interest | All other drugs | Total |
----------------------------------+------------------+-----------------+-------|
Adverse drug reaction of interest | A | B | A + B |
----------------------------------+------------------+-----------------+-------|
All other adverse drug reactions | C | D | C + D |
----------------------------------+------------------+-----------------+-------|
Total | A+C | B+D |A+B+C+D|
-------------------------------------------------------------------------------+
"""
# print """
#+-----------------------------------+------------------+-------------------------+
#| | Drug of interest | All other drugs | Total |
#+-----------------------------------+------------------+-----------------+-------|
#| Adverse drug reaction of interest | A | B | A + B |
#+-----------------------------------+------------------+-----------------+-------|
#| All other adverse drug reactions | C | D | C + D |
#+-----------------------------------+------------------+-----------------+-------|
#| Total | A+C | B+D |A+B+C+D|
#+-----------------------------------+------------------+-----------------+-------+
#"""
drug = drugs.standardize([drug])[0]
event = event.upper()
A = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = \"%s\" AND PT = \"%s\"" % (drug, event)
B = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME <> \"%s\" AND PT = \"%s\"" % (drug, event)
C = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME = \"%s\" AND PT <> \"%s\"" % (drug, event)
D = "SELECT SUM(COUNT) FROM temp.DRUG_EVENT_COUNT WHERE DRUGNAME <> \"%s\" AND PT <> \"%s\"" % (drug, event)
A = self.conn.execute(A).fetchone()[0]
B = self.conn.execute(B).fetchone()[0]
C = self.conn.execute(C).fetchone()[0]
D = self.conn.execute(D).fetchone()[0]
return {"A":A, "B":B, "C":C, "D":D, "drug":drug, "event":event}
#!/usr/bin/python
import drugstandards as drugs
import sqlite3
import csv
# Set the minmal Jario-Winkler similarity required for two strings to be considered a match.
THRESH = 0.9
# Establish connection to faers database which was created by running ./import_faers_data.sh
conn = sqlite3.connect("faers.db")
# Get unique drug names from faers.db to standardize.
faers = [i[0] for i in conn.execute("SELECT DISTINCT(drugname) FROM drug").fetchall() if i[0] != None]
# Standardize drug names form above.
stand = drugs.standardize(faers, thresh=THRESH)
pairs = [ (faers[i], stand[i]) for i in range(len(stand)) if stand[i] != None and faers[i] != stand[i]]
print "Adding %d records to drugmap" % len(pairs)
# Add DRUG_MAP table.
conn.execute('CREATE TABLE drugmap (original TEXT, replacement TEXT)')
conn.executemany('INSERT INTO drugmap (original, replacement) VALUES (?, ?)', (pairs))
conn.execute('CREATE INDEX drugmap_idx ON drugmap (original, replacement)')
conn.commit()
