Searching the literature

Searching the biological literature

Pubmed / Entrez

Essentially all biologically relevant literature is indexed by the National Library of Medicine, and linked to the Pubmed database. The entry point of choice for a search is Entrez, which, in addition to searching in Pubmed also searches a wide variety of other databases, including Genbank.

Sophisticated searches are possible in Pubmed, using the advanced search page. See the Pubmed Tutorial.

Some literature is missing from Pubmed, notably agricultural data (See Agricola for this) and literature pre-1960 or so. The coverage varies with the journal indexed, so a given article may not be present. Other indices, such as Web of Science are more useful for very old references.

Web of Science

The Science Citation Index or Web of Science is an extremely useful tool for searching forward in time from a relevant article. The bibliography of a research paper provides excellent links backward in time to relevant prior work; the science citation index provides a similar mechanism forward in time. Each article, when published, is mined for its references. SCI allows users to search for future articles which reference the article in question. Access is limited, as this is a commercial service. MIT users can access SCI from the Vera home page, A Web of Science tutorial is available here. Briefly, you almost always want a "Cited Reference Search." For the article you are searching from, enter the author name, first initial, and * as a wild card for any middle initial. Skip the journal, but enter the year of publication and hit Search. Locate the article you are searching from. In general there may be many instances with slightly varying references, typically due to errors and inconsistencies in the bibliographies. Press "Finish Search" to identify the articles which reference your article.

Science Citation Index is also useful in finding titles and references to articles which are not otherwise indexed, including Ph.D. theses, book chapters, patents, and web pages. Pubmed does not index these, while SCI does, since it indexes all references from indexed articles.

Navigating the journal pages

Pubmed indexes several types of books as if they were journals. Libraries are inconsistent about the corresponding cataloging. Notable examples are the Annual Reviews books, which are indexed as journals in Pubmed but shelved and cataloged as books at MIT (but not at Countway). An even more annoying example is the Methods in Molecular Biology series published by Humana Press. Here, Pubmed indexes the series as a journal, failing to give you the book title or author. The books are cataloged by topic and author at both MIT and Harvard. Methods in Enzymology is indexed as a journal in Pubmed, not available online, and is kept in the reference section at the MIT Science Library (or, for many important volumes, on TK's bookshelves).

Using the Libraries

University libraries house paper collections of many important journals. The MIT collections can be searched with Barton, the online catalog. Search Barton under "Journals" for information about the paper collections accessible at MIT libraries. Almost all biology related materials are in the basement of the Science Library at Hayden. Searches in Barton often fail when using the standard Pubmed journal abbreviation. Barton is poor in identifying the accessible online journals for MIT researchers. The Vera index of electronic resources captures some, but not all, of the online resources. Some of these resources can only be accessed by linking through the Vera page. Vera recognizes the Pubmed journal abbreviations much more readily than Barton. In some cases, MIT has access to journals which are neither indexed in Barton nor listed in Vera. The only way of knowing is to access the journal web site directly and see if you have full access to the articles.

Sometimes, but not always, there are links from the NCBI Pubmed site directly to the publisher's web site. Following these links often allows access to the article, but NB the point made earlier about required linking through Vera. If you think this is a mess and confusing, you are right.

Most journals have current content accessible on line. MIT may or may not have access to it. MIT access changes with time, and just because you have access today does not mean you will tomorrow. Elsevier journals, in particular, are essentially all accessible this year, through special arrangement. Next year they will not be. Content older than 1995 or so is normally inaccessible online, with important exceptions. The ASM is putting back issues of all of its journals online, including J Bact. Science and PNAS have back issues online. The Science back issues are (ironically) not accessible through the Science home page; you need to follow the JSTOR link in Vera to find older articles in Science. Nature is particularly annoying (remember this when you publish!!). They keep the last 5 years accessible online, and, if MIT pays enough, they can access older issues online. Currently MIT pays for the last 10 years.

The Patent Literature

The US Patent Office maintains an excellent search site for patents and published patent applicatons. One thing to know about this is that inventor names are listed in full, and that the wild-card character for the search is a dollar sign, rather than a star. Searches such as in/knight-t$ will find all of the inventions for Tom Knight, e.g.

The result of the USPTO search will be HTML versions of the patent or patent application. The diagrams and scanned images are not easily available from USPTO. They can, however, be downloaded for free from the web site Free Patents Online or Pat2PDF as a PDF document. European patents are also available on the first site.

There are no copyrights on patent documents. They can be freely uploaded and distributed in any form.

Automatic Updates

Permanent searches can be uploaded to the My NCBI web site, and run automatically, with results emailed periodically. This is an effective way of keeping up to date with recent literature in narrowly focused areas. Set up the search after logging into the My NCBI site (free registration).

Another service available is the free Pubcrawler which also searches Genbank files for new sequence data.

Online Techniques Forums

There are several online forums addressing the general topic of molecular biology technique. Protocol Online captures a set of online protocols, and an active community forum on techniques. Biotechniques, a free print journal of techniques, runs a less active forum. It is worthwhile subscribing to Biotechniques, if only to gain access to their online archives of articles, inaccessible in other ways. There is a very active Chinese forum here.

Check out Siphs and Connotea.