NCBI and pubmed data mining

Question

[1769mkc]

So I'm using this "RISmed" library to do some query of my gene or protein of interest and the output comes with pubmed ID basically, but most of the times it consist of non-specific hits as well which are not my interest.As I can only see the pubmed ID so i have to manually put those returned ID and search them in NCBI to see if the paper is of my interest or not.

So my question : Is there a way to to return the abstract of the paper or summary sort of along with its pumed ID , which can be implemented in R?

If anyone can help it would be really great..

Answer 1

[Jean-Karim Heriche]

Use the getAbstracts() function from the PubMedWordCloud package. Assuming you got a list of PMIDs in the pmids variable, the following should do it:

   library(PubMedWordCloud)
   abstracts<-getAbstracts(pmids)

Note also that RISmed has an AbstractText method (example here) but it seems more complicated to use.

[1769mkc]

thank you , i will try both of them

[1769mkc]

the error Im getting is this

"Unknown IO errorfailed to load external entity "http://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=26018226&rettype=abstract"

[GenoMax]

NCBI now uses https-only connections. Perhaps that is a thing to check into.

[1769mkc]

okay so how do i include that in my R query ?any suggestion

[Jean-Karim Heriche]

Good point about https. I submitted a request to the package maintainer. An alternative could be to use the RefManageR package with something like:

refs <- ReadPubMed(query=pmids, database = "PubMed")
abstracts <- refs$abstract