But today I found a paper that has a PMID (PMID: 13774225) and a corresponding DOI ( 10.1126/science.132.3434.1099), still the DOI was not retrievable through Efetch.
Does anyone know a mapping service for PMID's to DOI's?
Michael, I don't believe there is any requirement for everything that gets assigned a PMID to have a DOI. The identifiers are in no way related or dependent on each other. DOI's are issued by the journals, so if a journal doesn't issue them, or back date them, no such mapping will exist. Consequently the best way to map ascertain this would be to use something like the above anyway..
Daniel, its inconsequential that there's not a perfect PMID to DOI mapping. My question was more along the lines of a downloadable database of the known relationships - I wouldn't want to send 19 millions queries to eutils!
According to [?]Pubmed help[?] you were right. It should be where you looked for it. The article you were looking for is over 50 years old, so it must have been entered into Pubmed manually. I think they just forgot to add the DOI and you should send them the info. They will still probably have the best mapping service between PMID and DOI.
Related question: Anyone knows of a database of PMID's to DOI's? In the Medline XML not every PMID has a DOI
Michael, I don't believe there is any requirement for everything that gets assigned a PMID to have a DOI. The identifiers are in no way related or dependent on each other. DOI's are issued by the journals, so if a journal doesn't issue them, or back date them, no such mapping will exist. Consequently the best way to map ascertain this would be to use something like the above anyway..
Daniel, its inconsequential that there's not a perfect PMID to DOI mapping. My question was more along the lines of a downloadable database of the known relationships - I wouldn't want to send 19 millions queries to eutils!
I know of this downloadable file made by the Bioassist group of NBIC: http://bet1.nbiceng.net/temp/pmid2doi.sql.gz