As long as informatics on biologists counts as "bioinformatics", I hope this question can find an answer here...
I am trying to build a co-author graph based on institutional affiliations. I have a list of authors (or even a list of pubmed IDs) and thought I'd simply go to Pubmed and extract affiliation data from their xml return. BUT, to my surprise, they only provide the affiliation for the first author. What am I missing? What other sources do folks use for author affiliation data?
2 answers
You're right, Pubmed only contains the affiliation for the 1st author. You can always get this information from the XML dump of the articles in PMC:
e.g: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pmc&id=114965
<contrib contrib-type="author">
<name>
<surname>Poncet</surname>
<given-names>Didier</given-names>
</name>
<xref ref-type="aff" rid="N0x97cf288.0x8bf6180">2</xref>
</contrib>
(...)
<aff id="N0x97cf288.0x8bf6180"> Laboratoire de Virologie, Hôpital Armand Trousseau (EA 2391, UFR Saint-Antoine), Paris,<sup>1</sup> and Unité INRA 1144 Virologie Moléculaire et Cellulaire, 78852 Jouy-en-Josas,<sup>2</sup> France</aff>
but beware: you cannot dump a large number of articles from PMC.
update (27 Sept) : the pubmed DTD with change next year and you'll get one affiliation per author: http://eutils.ncbi.nlm.nih.gov/entrez/query/DTD/nlmmedlinecitationset_140101.dtd
<Author ValidYN="Y">
<LastName>Billings</LastName>
<ForeName>J Andrew</ForeName>
<Initials>JA</Initials>
<Affiliation> Palliative Care Service, Massachusetts General Hospital and the Harvard Medical School Center for Palliative Care, Boston, Massachusetts, USA.</Affiliation>
</Author>
Log in to answer this question.
This doesn't help solve your problem, but NCBI recently announced that they will be changing this practice and will index all author affiliations submitted by publishers.