Controlling the flood of biological information
Immunology Today. (1998)18:557-558.
Human biology is enormously complex, as is the immune response. Creating important new information depends on the increasingly difficult task of identifying and integrating what is currently known. The urgency of addressing these issues has increased dramatically due to the progress of the human genome project towards defining the estimated 100,000 genes, as well as the advent of other 'high-throughput' biological techniques. It is essential to find organizing strategies to take advantage of this information, rather than make researchers suffer trying to find information. On the basis of past experience, together with discussions with literally hundreds of biologists over the past several years, we have evolved a strategy.
The resource we are building is called PROW: Protein Reviews on the Web. Its founding principles are summarized in Box 1. PROW has completed most of its first objective: making available on the World Wide Web (WWW) short authoritative 'guides' on the approximately 200 human CD cell-surface molecules (Box 2). This PROW 'pilot project' has built upon the International Workshops on Human Leukocyte Differentiation Antigens (IWHLDA). During the past 17 years, IWHLDA has generated a tradition of cooperative creation among scientists in providing authoritative information on human cell-surface proteins. PROW aspires to expand this spirit to the creative assembly of information on these and other human proteins.
PROW guides are short, ~1000 word, structured reviews that collate all available information on the molecule and, importantly, are all peer-reviewed. To see the layout of a PROW guide goto CD127.
The format differs from conventional published reviews in four critical ways. First, PROW guides have approximately 20 standardized 'categories' of information (biochemical function, ligands, etc.) in order to give readers a standard layout to expect, and to encourage authors to be systematic in their approach to describing the molecule. Second, the facts in each PROW category are single sentence/phrase entries, rather than paragraphs. This encourages brevity, and facilitates revision. Third, the format is optimized for WWW presentation, taking advantage of hypertext links to provide details beyond the scope of the short review format. For example, it links to outstanding resources such as SwissProt, Mendelian Inheritance in Man and Medline/Pubmed. Finally, in each category, readers have the opportunity to add comments (with citations), thereby creating an accompanying 'forum' page.
Although PROW is a shoestring operation, encouraging progress has been made. The biggest source of satisfaction has been the cooperation of the biologists writing and reviewing the guides and the quality of the contributions they have made. One delightful outcome has been the substantial additions by reviewers, and the resulting e-mail dialog between authors and reviewers who were often previously unacquainted. We believe that support by biologists will be the most critical determinant of the success or failure of PROW. Therefore, we have designed the process of writing and reviewing the guides to be as painless as possible; for example, all correspondence is handled by e-mail, a text template of categories is provided to the author, the author does not have to worry about special formatting such as references (instead it is done semi-automatically by PROW), and links to resources such as SwissProt are added by PROW not the authors. We expect PROW guides to be cited as references and thus provide some additional payback for the effort biologists expend on this project.
In thinking about PROW, we strive to maintain both a short-term (this year) and a long-term (10 years from now) perspective. We want to develop strategies that can evolve progressively to make learning biology, and doing biological research, more fun and productive. Although PROW is envisaged as a resource on human genes/proteins, the insights from research on homologous molecules in other species (mouse, fly, yeast, worm, etc.) will be reflected either indirectly or directly in PROW. In the short term, we will add information on molecular families. Gradually, with the involvement of colleagues around the world, we expect to expand to other molecular families and to incorporate the concepts of associated molecules and biochemical pathways. The visionary concept of an 'electronic cell', which models cell behavior, is a long way from implementation. However, PROW is one of the evolutionary processes that are necessary to progress towards that goal.
Box 1. The founding principles of PROW
PROW will organize human biological information around the gene, which we believe to be the most 'durable' and powerful organizing principle. Authoritative information is contributed to PROW by the real experts in the subject matter, (the biologists), thus validating the information Information from PROW is available to the whole biological community, not restricted to the wealthy few who can afford high-priced proprietary services PROW will be international and not 'belong' to a particular country or special interest group PROW will creatively exploit computerized stategies for authoring, maintenance and distribution of information, so that it will be cost-efficient and scalable from an initial pilot project to one that includes all of the estimated 1000,000 genes
Box 2. PROW on the Web
Prow guides to the CD antigens can be found on the World Wide Web at ncbi.nlm.nih.gov
By its nature, the PROW resource has been the beneficiary of help from many people and organizations (http://www.ncbi.nlm.nih.gov/PROW/rem/credits). Special thanks to the National Cancer Institute (NCI), the National Center for Biotechnology Information (NCBI), the International Workshop on Human Leukocyte Differentiation Antigens (IWHLDA), Garland Press and the >300 experts who have written and reviewed the CD guides.
Stephen Shaw (sshaw@nih.gov), Lisa Turni and Kenneth Katz are at the National Cancer Institute, Bethesda, MD 20892, USA.
Jim |
