GoPubMed: the knowledge-based PubMed search

GoPubMed is a search engine that provides PubMed results "sorted into meaningful categories," as well as the ability to narrow those results in ways that are meaningful to the searcher, providing relevant citations without the need to comb through an overwhelming results list.

GoPubMed uses the Gene Ontology (GO) and the National Library of Medicine's Medical Subject Headings (MeSH) as knowledge structures to determine how to sort these results, and claims that the result categories will "in many cases ... already represent the answer to your question," which is certainly a different way of discovering information in PubMed's enormous - again, sometimes overwhelming - database.

The example GoPubMed provides on the FAQ page is a search for the protein Rab5, which turns up a list of top related terms as well as categories this term appears in in the Knowledge Base, both of which might lead to a more expansive perspective on the original search question.  As promised, navigating through the drop-down menus below each category does provide a wealth of information, which might both answer the question that prompted the initial search as well as provide new terms with which to build a more in-depth search. The ability to narrow from here - to include specific resulting terms in a narrower search, or to exclude ones that are way off - leads to a custom-built results set.

As you might expect, search results can also be narrowed by location, publication or publication date; unfortunately, they can't yet be narrowed by what a particular institution has access to, so if you're searching GoPubMed from MGH for articles rather than the answer to a basic question, the best option would be to find citations via this search engine and then click over to our website and search them from our link to PubMed, which can be limited to resources we provide access to.  

If you were previously a MeshPubMed user, this service has now been integrated into GoPubMed.