Add PMID or Accession Number in enriched output?

[globbestael]

What

Should DedupEndnote copy the PMID of a duplicate record to the record kept?

OR: should all IDs in the AN field be deduplicated and added to the record kept?

Background

EndNote uses the PMID in:

• ??? check for update
• ??? deduplication (or is it PMCID?)
• checking for retractions

Other questions:

• in what fields does EndNote look for a PMID?
• what rules are used by EndNote for identifying a PMID. E.g. is any number in the AN field a potential PMID / used by EndNote when checking for updates / retractions / ...
• is AN field a repeatable field?
• what happens if PMID is preceded in EndNote by "PMID " / "PMID: " / "PMID:"
  • Compare the last one with "WOS:000505957800099" for a Web of Science record. In the current test files only WoS IDs use a prefix ("WOS:")
  • searching in PubMed on "pmid:17714497" finds the record with PMID 17714497. Advanced search shows that PubMed is searching "17714497[UID]"
  • searching in PubMed on "17714497" finds the record with PMID 17714497. Advanced search shows that PubMed is searching "Schema: all" (meaning all fields?)
  • searching in PubMed on "pmid:23" finds the record with PMID 23, but "pmid:023" finds the record with PMID 19, because the ID is read as an octal number!
  • searching in PubMed on "WOS:A1997WA65000036" actually searches on "WOS" and "A1997WA65000036", and the first search items had results
  • searching in PubMed on "pmid:73091900004" (this PMID does not exist), shows in Advanced search 2 records, the first one with "Schema: all", the second one without, and both as "73091900004[pmid]" (NO translation of "pmid" to "UID").
• What rules should DedupEndNote use to identify an ID as a PMID?
  • If DedupEndNote prefixes PMIDs with "PMID:" it better be sure that the ID is indeed a PMID and not the ID of another database.
• PMIDs may be available in other data, e.g. as a parameter in the URL : http://www.ezproxy.is.ed.ac.uk/login?url=http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed11&AN=352031592
  http://openurl.ac.uk/athens:_edu//lfp/LinkFinderPlus/Display?sid=OVID:Embase&id=pmid:18572984&id=&issn=0160-6689&isbn=&volume=69&issue=7&spage=1092&pages=1092-1098&date=2008&title=Journal+of+Clinical+Psychiatry&atitle=A+preliminary+study+of+fMRI-guided+rTMS+in+the+treatment+of+generalized+anxiety+disorder&aulast=Bystritsky&pid=
  Extracting PMIDs is not a subtask of deduplication, but could be useful.
  • But then we should also look at embedded DOIs, e.g. http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=21833673
    http://ZL9EQ5LQ7V.search.serialssolutions.com/?sid=OVID:medline&id=pmid:21833673&id=doi:10.1007%2Fs12013-011-9258-5&issn=1085-9195&isbn=&volume=62&issue=1&spage=55&pages=55-8&date=2012&title=Cell+Biochemistry+%26+Biophysics&atitle=Treatment+with+Huperzine+A+improves+cognition+in+vascular+dementia+patients.&aulast=Xu&pid=%3Cauthor%3EXu+ZQ%3BLiang+XM%3BJuan-Wu%3BZhang+YF%3BZhu+CX%3BJiang+XJ%3C%2Fauthor%3E%3CAN%3E21833673%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E