Skip to content

Instantly share code, notes, and snippets.

@harej

harej/super_pubmed.py

Created September 25, 2016 03:27
Show Gist options
  • Save harej/0ae3b77db7c1114cec3dbb0b8f1664cc to your computer and use it in GitHub Desktop.
Save harej/0ae3b77db7c1114cec3dbb0b8f1664cc to your computer and use it in GitHub Desktop.
Download ZIP
Raw
super_pubmed.py
import html
import requests
import threading
class AskPubMed(threading.Thread):
def __init__ (self, threadID, name, packages):
threading.Thread.__init__(self)
self.threadID = threadID
self.name = name
self.packages = packages
def run(self):
esummary_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&retmode=json&tool=wikidata_worker&[email protected]&id="
idconv_url = "https://www.ncbi.nlm.nih.gov/pmc/utils/idconv/v1.0/?format=json&tool=wikidata_worker&[email protected]&ids="
months = {
"Jan": "01",
"Feb": "02",
"Mar": "03",
"Apr": "04",
"May": "05",
"Jun": "06",
"Jul": "07",
"Aug": "08",
"Sep": "09",
"Oct": "10",
"Nov": "11",
"Dec": "12"
}
for package in self.packages:
bunch_of_numbers = ""
for pmid in package:
bunch_of_numbers += pmid + ","
bunch_of_numbers = bunch_of_numbers[:-1] # Remove trailing comma
summary_retriever = requests.get(esummary_url + bunch_of_numbers)
if summary_retriever.status_code != 200:
continue
# Now processing the bibliographic metadata from our summary retriever query...
summary_retriever_json = summary_retriever.json()
if "result" in summary_retriever_json:
for _, pmid_blob in summary_retriever_json["result"].items():
if _ == "uids":
continue
pmid = pmid_blob["uid"]
# First: The basics
output_string = "CREATE\n"
output_string += "LAST\tP698\t\"" + pmid + "\"\tS248\tQ180686\n"
output_string += "LAST\tP31\tQ13442814\tS248\tQ180686\n"
output_string += "LAST\tDen\t\"" + "scientific article" + "\"\n"
# Are there other IDs we can add?
doi = None # if there is a DOI, this value will be overridden
if "articleids" in pmid_blob:
for identifier in pmid_blob["articleids"]:
if identifier["idtype"] == "doi":
doi = identifier["value"] # We want the DOI for later
output_string += "LAST\tP356\t\"" + identifier["value"] + "\"\tS248\tQ180686\n"
elif identifier["idtype"] == "pmc":
pmcid = identifier["value"].replace("PMC", "")
output_string += "LAST\tP932\t\"" + pmcid + "\"\tS248\tQ180686\n"
# Title
if "title" in pmid_blob:
t = html.unescape(pmid_blob["title"])
if t != "":
output_string += "LAST\tLen\t\"" + t + "\"\n"
output_string += "LAST\tP1476\ten:\"" + t + "\"\tS248\tQ180686\n"
# Publication date
if "pubdate" in pmid_blob:
pubdate = None
pubdate_raw = pmid_blob["pubdate"].split(" ") # 2016 Aug 1
if len(pubdate_raw) > 1:
if pubdate_raw[1] in months:
m = months[pubdate_raw[1]]
else:
m = "00"
if len(pubdate_raw) == 3: # Precision to the day
pubdate = "+{0}-{1}-{2}T00:00:00Z/11".format(pubdate_raw[0], m, pubdate_raw[2].zfill(2))
elif len(pubdate_raw) == 2: # Precision to the month
pubdate = "+{0}-{1}-00T00:00:00Z/10".format(pubdate_raw[0], m)
elif len(pubdate_raw) == 1: # Precision to the year
pubdate = "+{0}-00-00T00:00:00Z/9".format(pubdate_raw[0])
if pubdate != None:
output_string += "LAST\tP577\t" + pubdate + "\tS248\tQ180686\n"
# Published in
if "issn" in pmid_blob:
issn_query_url = "https://query.wikidata.org/sparql?format=json&query=select%20%3Fi%20%3Fissn%20where%20%7B%20%3Fi%20wdt%3AP236%20%22{0}%22%20%7D"
issn_query = requests.get(issn_query_url.format(pmid_blob["issn"])).json()
issn_results = issn_query["results"]["bindings"]
if len(issn_results) == 1: # We want no ambiguity here
journal = issn_results[0]["i"]["value"].replace("http://www.wikidata.org/entity/", "")
output_string += "LAST\tP1433\t" + journal + "\tS248\tQ180686\n"
# Volume
if "volume" in pmid_blob:
if pmid_blob["volume"] != "":
output_string += "LAST\tP478\t\"" + pmid_blob["volume"] + "\"\tS248\tQ180686\n"
# Issue
if "issue" in pmid_blob:
if pmid_blob["issue"] != "":
output_string += "LAST\tP433\t\"" + pmid_blob["issue"] + "\"\tS248\tQ180686\n"
# Pages
if "pages" in pmid_blob:
if pmid_blob["pages"] != "":
output_string += "LAST\tP304\t\"" + pmid_blob["pages"] + "\"\tS248\tQ180686\n"
# Original language
if "lang" in pmid_blob:
for langcode in pmid_blob["lang"]:
if langcode == "eng":
output_string += "LAST\tP364\tQ1860\tS248\tQ180686\n"
break
# Authors
authors_not_done = True # set to False if authors are successfully extracted via Crossref
if doi != None:
crossref = requests.get("https://dx.doi.org/" + doi, headers={"Accept": "application/json"})
if crossref.status_code == 200:
try:
crossref_json = crossref.json()
if "author" in crossref_json:
authors_not_done = False
author_counter = 0
for author in crossref_json["author"]:
author_counter += 1
a = ""
if "family" in author:
a = author["family"]
if "given" in author:
a = author["given"] + " " + a
output_string += "LAST\tP2093\t\"" + a + "\"\tP1545\t\"" + str(author_counter) + "\"\tS248\tQ5188229\n"
except ValueError:
pass
if "authors" in pmid_blob and authors_not_done == True:
author_counter = 0
for author in pmid_blob["authors"]:
if author["authtype"] == "Author":
author_counter += 1
output_string += "LAST\tP2093\t\"" + author["name"] + "\"\tP1545\t\"" + str(author_counter) + "\"\tS248\tQ180686\n"
output_string = output_string[:-1]
print(output_string)
def main(seed_url):
seed = requests.get(seed_url).json()
full_pmid_list = [x for x in seed["esearchresult"]["idlist"]]
wikidata = requests.get("https://query.wikidata.org/sparql?format=json&query=select%20%3Fp%20where%20%7B%20%3Fi%20wdt%3AP698%20%3Fp%20%7D").json()
wikidata_pmid_list = [x["p"]["value"] for x in wikidata["results"]["bindings"]]
pmid_list = list(set(full_pmid_list) - set(wikidata_pmid_list))
# A list of 200 IDs makes a package. These collectively are the "packages".
# The package of all these packages is the "Package of Packages".
packages = [pmid_list[x:x+200] for x in range(0, len(pmid_list), 200)]
package_of_packages = [packages[x:x+225] for x in range(0, len(packages), 225)]
thread_counter = 0
for packages in package_of_packages:
thread = AskPubMed(thread_counter, "thread-" + str(thread_counter), packages)
thread_counter += 1
thread.start()
if __name__ == "__main__":
main("https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?retmode=json&[email protected]&tool=wikidata_worker&db=pubmed&term=review[filter]%20free%20full%20text[filter]&reldate=1850&datetype=edat&retmax=180000")
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment