Invalid PMIDs in postgres and WormBase
In the course of trying to update the PaperType tag in the postgres and WormBase data models, we came across 13 WBPapers with invalid PMIDs.
Here is a breakdown of why those PMIDs are now invalid, if known, and suggested actions:
Corrigenda and Errata
Corrigenda and errata are not valid entries in PubMed, but are sometimes mistakenly sent to PubMed by the journals. Until the correction is manually curated and removed from the database, it will have a PMID. Once removed, that corresponding PMID is no longer valid. Upshot: don't take corrections from PubMed. This raises the issue of how many corrections are out there that we don't know about, but finding and updating all of them would be a considerable task.
PMID:16240344 WBPaper00026911 Correction for: PMID:16130113 WBPaper00026758
PMID:18548071 WBPaper00031950 Correction for: PMID:18185515 WBPaper00031383
PMID:18841162 WBPaper00032248 Correction for: PMID:18701922 WBPaper00032104
PMID:19435507 WBPaper00033151 Correction for: PMID:18211672 WBPaper00031447
Suggested action: Since the PDFs associated with the corrections does actually contain the corrected material, make the PMID for each correction invalid in postgres and change the corresponding PaperType to Correction.
Papers either not included in or apparently removed from PubMed
These are papers that must have had a valid PMID in PubMed, but for some reason were removed from the PubMed database. In one case, it looks like the paper is actually a meeting abstract, so maybe that's why it was removed. In the other, it's not clear why the paper is not included.
Suggested action: Make the PMID invalid in postgres and change the corresponding PaperType to Meeting_Abstract and Article, respectively.
Paper is valid but has a new PMID
In a few cases, the PMID associated with a WBPaper was invalid, having been replaced by a new PMID in PubMed. These may be cases where duplicate entries were sent to PubMed or for some other reason the original PMID was updated. The new PMID is NOT associated with any other WBPaper ID in postgres.
PMID:16636786 WBPaper00027350 Current PMID:16758254
PMID:16652241 WBPaper00027366 Current PMID:16752214
PMID:18677322 WBPaper00032079 Current PMID:19160519
PMID:18711361 WBPaper00032109 Current PMID:19160516
Suggested action: Make the old PMID invalid in postgres and add the new PMID. Change the PaperType to Article for each of these.
Paper is a duplicate of another WBPaper with a different PMID
These are cases where there are two entries in postgres for a paper and it looks like the old entry was never made invalid. I don't know why there were duplicate entries in PubMed.
PMID:16641367 WBPaper00027353 Current PMID:16801682 WBPaper00027709
PMID:17154166 WBPaper00028879 Current PMID:17443782 WBPaper00029300
PMID:17154292 WBPaper00028880 Current PMID:17443777 WBPaper00029299
Suggested action: Merge the two WBPapers and make the first PMID and WBPaper IDs invalid.