[BiO BB] Programmatic access to pubmed

[Dr. Christoph Gille]

In our team we reconstruct metabolic networks and therefore we need to screen
hundreds of pubmed references to find evidences for certain
metabolic reactions in the literature.

We use this simple Java-application:
http://www.bioinformatics.org/strap/strap.php?pubmed=t

We proceed as follows:
We make a list of Pmid-numbers.

We move the mouse over the list and then we go for a coffee to
give the system sufficient time to cache the Abstracts.

Then we move again the mouse over the list of PMID numbers.  This
time we observe color text high-lightings (defined by
Ctrl-F "Find") to appear in the abstract panel.

If abstract is too big, then the high-lightings may appear in the
vertical scroll-bar of the text panel.


The system might automatically identify full text links by following the links provided by NCBI.
In all cases PDF can be manually associated.

In this case this visual text-mining can be performed for full text.



http://www.bioinformatics.org/strap/strap.php?pubmed=t

Appart from this, professional Literature managers such as Jabref or Mendeley are able to load lists of pubmed abstracts.