27 January: Primary Literature

Primary literature reports the results of research, and is mainly of interest to other workers in particular subdisciplines --people who are part of the ongoing conversations of science. Your task is to (1) locate and (2) assimilate (comprehend, interconnect) the research literature relevant to your topic: in short to eavesdrop on a conversation and learn enough of its terminology and findings and directions to be able to summarize it succinctly.

Last week's activities with the secondary/review literature were designed to help you find on-ramps to ongoing conversations, and the next few weeks will build on that foundation. You can probably (in fact, almost certainly) use the review article(s) you found to direct your search for the primary sources you need: you can ask the questions

In short, in the context of your topic, you're in search of answers to these questions:
What are the interesting open questions, and how are people going about answering them?

Before we get to the specific tools for searching for primary articles, a bit of continuation of last week's examination of Review literature:

I suggested a strategy for finding Review articles, by looking in the first paragraphs of a Research article to see what the author(s) said about summaries of the research subfield in which they are working. A case in point: Science has this article, chosen more or less at random: Phosphorus Limitation of Coastal Ecosystem Processes (P. V. Sundareshwar et al. Volume 299, Number 5606, Issue of 24 Jan 2003, pp. 563-565). In the first paragraph, the authors cite
1. P. M. Vitousek, et al., Ecol. Appl. 7, 737 (1997) [ISI]
...but that information isn't enough for us to work with if we need to decide if Vitousek et al. really is a Review article, or even if we want to retrieve it because their brief description ("It has been well established that nitrogen is a major nutrient that limits primary production in the coastal zone (1)...") suggests that it's right at the center of whatever we are seeking.

So what's the problem here, and how do we solve it? How do we find and retrieve this Vitousek et al. 1997?.

We are looking at Science's citation format: every journal has one, and there is no single "right way" to cite that fits every journal. Sooooo.... you have to learn to cope with the conventions and the workarounds.
  1. Abbreviations can be decoded in a number of ways. There's a book in the Science Library's Reference collection that's a pretty complete source for abbreviations"
    TITLE        BIOSIS serial sources.
IMPRINT      Philadelphia, PA : BIOSIS, 1995-
CALL NO.     SCI REF Z5321.B56.
    There are online sources for abbreviations: Cal Tech has an online page of ISI Journal Abbreviations that seems to work pretty well... and there are others linked from the Science Library's Biology Dept page (Journal abbreviations). This one turns out to be Ecological Applications, which an Annie search tells us is available in 3 formats: in paper, from JSTOR (1991-1998), and from the publisher (1997-)

  2. But what about that [ISI]?
    ISI is Institute for Scientific Information, the vendor of Web of Science --which you used for its 'Review' feature. We need to explore what it is that ISI does, to get some idea of why we should care, and what trust we should put in its search capabilities. Knowing that Science identifies an article with [ISI] is actually useful, and the [ISI] link will take you directly to the Web of Science record for the article.

    You can also get to the article's record by going to the Web of Science search interface:

    1. Go to isiknowledge.com
    2. Once in Web of Science, choose 'Advanced Search', then choose Cited Ref Search
    3. In this case we type Vitousek p* and 1997 and click Lookup. We get

      1	VITOUSEK PM	ECOL APPL	1997	7	373	 	 
1	VITOUSEK PM	ECOL APPL	1997	7	735	 	 
1	VITOUSEK PM	ECOL APPL	1997	7	736	 	 
481	VITOUSEK PM	ECOL APPL	1997	7	737	 	
1	VITOUSEK PM	ECOL APPL	1997	7	7376	 	 
1	VITOUSEK PM	ECOLOGICAL APPL	1997	54	119	 	 
1	VITOUSEK PM	ECOLOGICAL APPL	1997	7	739
      ...and what does that mean? ==> 487 subsequent authors have cited this article, the ISI record for which says:

      Human alteration of the global nitrogen cycle: Sources and consequences
      Vitousek PM, Aber JD, Howarth RW, Likens GE, Matson PA, Schindler DW, Schlesinger WH, Tilman DG
      ECOLOGICAL APPLICATIONS
      7 (3): 737-750 AUG 1997

    Sort of a long way 'round to get the answer, but it worked. So now we have the information we need to put this reference into whatever format we might need to use, as defined by the journal for which we are writing, or the professor's preferences if it's to go into a bibliography.

Now, on to searching for Primary/Research articles. Last time we looked at the morphology of a Primary source --the main earmark is that a Primary article reports research findings, generally with data in some explicit form. And we observed that Primary articles have to fit themselves into existing literatures, by citing other articles, and eventually by being cited themselves. If nobody cites an article, its influence is negligible. If many people cite it, it may become a classic. Most are somewhere between those extremes, because most are read only by other specialists ...that's how science works.

ISI rates journals by their "impact": by how many citations the articles in them get from other articles, and the data are also used to compare universities (see Top Ten Universities in Biological Science Fields, 1993-97).

Web of Science is a very powerful basic search engine for the sciences, but it does have some limitations: it indexes about 8,000 journals which it considers the 'most important' ...and ignores many others. It makes no claim to being exhaustive. One might criticize the ISI databases as being limited to 'conventional science' --not such a bad thing, but not a recipe for discovery of things that are at the fringes, where some of the most exciting things may be happening.

Cambridge Scientific Abstracts tries to be more exhaustive, and the Biomedical suite includes Plant Science too.

And there are other specialized databases that are of importance to specific fields --thus, Agricola (USDA) is valuable for agriculture (obviously), but also for plants; and PubMed is an essential for the biomedical literatures; and (just a bit farther afield) PsychINFO deals with many branches of Psychology that overlap with Biology, and SciFinder Scholar is centered in Chemistry, but includes a lot of Biology and various physical sciences.

The ones you need to work with are Web of Science and Cambridge Scientific Abstracts, unless you're in Dr. I'Anson's, Dr. Eisen-Vandevelde's, or Dr. Wielgus's group (in which case you should also try PubMed).

An example: at the moment H5N1 is a very big deal in science news (see my log of the unfolding story), and one might want to know more of the state of the literature on the subject.

Annual Reviews gives me two hits:
Global Epidemiology of Influenza: Past and Present
N. J. Cox, K. Subbarao
Annual Review of Medicine. Volume 51, Page 407-421, Feb 2000

GENETICS OF INFLUENZA VIRUSES
David A. Steinhauer, John J. Skehel
Annual Review of Genetics. Volume 36, Page 305-332, Dec 2002

Cambridge Scientific Abstracts, limited to PY=Review, got 13 hits in January 2004, 12 of which looked significant (and some of which we have) ...Cox and Subbarao shows up, but not the other one (why?). This week there are 47...

...and Web of Science got 29 hits last year, and 11 more this year, which I've exported as .html

If I use Web of Science to do a shotgun search for H5N1 in all doc types, I got 233 hits in January 2004, and 306 this year... and sorting through those might give me the 'primary' articles I'd like to have... but another approach is to limit that 306 by adding the name of a prominent researcher. I'll use RG Webster... and a year ago I got 46 articles by adding him to the search as author. This year it's 56 --though it's interesting that he's not the firstauthor listed.
Here's one of them:

Emergence of multiple genotypes of H5N1 avian influenza viruses in Hong Kong SAR
Guan Y, Peiris JSM, Lipatov AS, Ellis TM, Dyrting KC, Krauss S, Zhang LJ, Webster RG, Shortridge KF
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
99 (13): 8950-8955 JUN 25 2002
...which has "materials and methods" and "results" sections, so I know it's a 'primary' article. It's been cited by 27 subsequent authors (but only 11 a year ago), one of which is

The influenza virus gene pool in a poultry market in South Central China
Liu M, He SQ, Walker D, Zhou NN, Perez DR, Mo B, Li F, Huang XT, Webster RG, Webby RJ
VIROLOGY 305 (2): 267-275 JAN 20 2003

The Guan et al. article's "Related Records" search pulls up 961 articles with varying degrees of overlap of bibliography (800+ a year ago), the first (strongest) of which is

Tumpey TM, Suarez DL, Perkins LEL, et al.
Characterization of a highly pathogenic H5N1 avian influenza a virus isolated from duck meat
J VIROL 76 (12): 6344-6355 JUN 2002

...so what I have here is a good start at an interlinked cluster of articles that bear upon this very important issue of the plasticity of the H5N1 virus...

Just to round things off, and because some of you have topics that are definitely 'bio-medical', let's also look quickly at PubMed, from the National Library of Medicine. There's a link to the search interface on the Biology page, just under Cambridge Scientific Abstracts.
If I do a search for 'h5n1'I get 242 hits (186 a year ago). This year the most recent articles are authored by Japanese, Thai, Chinese researchers --underlining the fact that science is global/international. And PubMed seems to be the most frequently updated, for this rapidly-developing subject area. Last year, the very first one was very fresh (January 17 2004 issue of Lancet), by our friend Robert G. Webster (who must be a very busy man just now...) and has a link to full text. Amazing, and it's because we HAVE a subscription to The Lancet (and PubMed 'knows' that). In many cases you can't get to the full text via the records, because we haven't paid for access, and that's what InterLibrary Loan is for.

Note that PubMed has a 'Related Articles' button... a very powerful feature.

===

BIOSIS Biology Browser