biomedical research

Based on studies of hypophosphatasia, which is a systemic skeletal disorder resulting from tissuenonspecific alkaline phosphatase (TNSALP) deficiency, TNSALP was suggested to be indispensable for bone mineralization. Recently, we demonstrated that there was a significant difference in bone mineral density (BMD) among haplotypes, which was lowest among TNSALP (787T [Tyr246Tyr]) homozygotes, highest among TNSALP (787T > C [Tyr246His]) homozygotes, and intermediate among heterozygotes. To analyze protein translated from the TNSALP gene 787T > C, we performed the biosynthesis of TNSALPs using TNSALP cDNA expression vectors. TNSALP (787T) and TNSALP (787T > C) were synthesized similarly as a high-mannose-type 66-kDa form, becoming an 80-kDa form. Expression of the human 787T > C TNSALP gene using the cultured mouse marrow stromal cell line ST2 demonstrated that the protein translated from 787T > C exhibited an ALP-specific activity similarly to that of 787T. Interestingly, the Km value for TNSALP in ST2 cells transfected with the 787T > C TNSALP gene was decreased significantly compared to that of cells carrying the 787T gene (P < 0.01). These results suggest that the significant difference in Km values between the proteins translated from 787T > C and 787T may contribute to regulatory effects on bone metabolism. Alkaline phosphatase (ALP; orthophosphoric monoester phospho-hydrolase, alkaline optimum, EC 3.1.3.1.) is classified into two types in most animals, excluding homonidae: tissue-nonspecific (liver/bone/ kidney; TNSALP) and intestinal types (29). In humans, there are at least four types of genetically differing isozymes: tissue-nonspecific, intestinal, placental, and germ cell types (11, 18, 29). The TNSALP gene (GenBank: NM_000478) is located on chromosome 1 and consists of 12 exons and 11 introns, with the coding sequence beginning in the second exon (29). TNSALP shows an approximately 50% homology with the other three isozymes (intestinal, placental, and germ cell). Their isozymes are tissue-specific and their genes are 90 ~ 98% homologous and clustered on chromosome 2 (11, 18). The core structures are largely conserved and exhibit the same metal ions and glycosylation sites in all mammalian ALPs. As a result of studies on cDNAs enAddress correspondence to: Masae Goseki-Sone, Ph. D. Division of Nutrition, Department of Food and Nutrition, Japan Women’s University, 2-8-1, Mejirodai, Bunkyo-ku, Tokyo 112-8681, Japan. Tel: +81-3-5981-3429, Fax: +81-3-5981-3117 E-mail: goseki@fc.jwu.ac.jp (M. Goseki-Sone) N. Sogabe et al. 214 sis of TNSALP (787T > C) using an in vitro transcription/translation system. TNSALP is a marker enzyme of osteoblastic differentiation, and the enzymatic activity of TNSALP in the cultured mouse marrow stromal cell line ST2 was elevated significantly by Pi starvation (8) or bone morphogenetic protein-2 (BMP-2) treatment (9). Therefore, we examined the effects of amino acid substitution in TNSALP (787T > C) on the catalytic property in ST2 cells. MATERIALS AND METHODS Construction and analysis of the mutant TNSALP cDNA expression plasmid. Normal TNSALP cDNA was obtained from human periodontal ligament cells (5) and inserted downstream of the cytomegalovirus gene promoter (pCMV) of the expression plasmid vector pcDNA3 (Invitrogen Corp., Carlsbad, CA, USA), as described previously (7). To create polymorphism (Tyr246His, 787T > C) expression vectors, site-directed mutagenesis was performed using a Mutagenesis Kit (Amersham Pharmacia Biotech), as described elsewhere (10). Metabolic labeling and immunoprecipitation. COS-1 cells were cultured in Dulbecco’s modified Eagle’s minimum essential medium (DMED) (Gibco, Invitrogen Corp.) supplemented with 10% (v/v) fetal calf serum. COS-1 cells (1.0–1.3 × 10 cells/35-mmdiameter dish) were transfected with 0.8–1.0 μg of each plasmid using Lipofectamine Plus (Invitrogen Corp.) according to the manufacturer’s protocol, as described previously (14). Expre[S][S] protein labeling mix ( > 1000 Ci/mM) was obtained from Dupont-New England Nuclear (Boston, MA, USA). [C]methylated proteins, enhanced chemiluminescence Western blotting detection reagent, peroxidase-conjugated donkey ant-rabbit IgG, and Protein A-Sepharose CL-4B were from Amarsham Pharmacia Biotech (Arlington Heights, IL, USA). For pulse-chase experiments, cells were preincubated for 0.5–1 h in methionine/cysteine-free DMED and labeled with 5–100 μCi of [S]methionine/cysteine for 0.5 h in the fresh methionine/cysteine-free MEM. After metabolic labeling, the medium was removed, and the cells were lysed in 0.5 mL of lysis buffer [1% (w/v) SDS in NaCl/Pi]. The lysates were subjected to immunoisolation, as described previously (14, 27). The immune complexes/Protein A beads were boiled in the presence or absence of 1% (v/v) 2-mercaptoethanol, and then analyzed by SDS/ PAGE [9% (w/v) gels], followed by fluorography coding ALP isozymes, it is known that the primary structure in the catalytic region is well-conserved in ALPs of humans, animals, and E. coli., suggesting that TNSALP plays an important role in active metabolism by hydrolyzing phospho-compounds, supplying free inorganic phosphate (Pi). The physiological roles of ALP are not been wellunderstood, but strong evidence is provided by the rare genetic disease hypophosphatasia (HPP). HPP is an inherited disorder characterized by a defect in skeletal mineralization caused by TNSALP deficiency (30). Various mutations in the TNSALP gene have been analyzed (2, 3, 6, 7, 12, 14, 15, 19, 23– 28). Elevated extracellular concentrations of inorganic pyrophosphate (PPi), phosphoethanolamine (PEA), and pyridoxal-5’-phosphate (PLP) have been observed in HPP (30). Osteoporosis is defined as a skeletal disorder characterized by compromised bone strength, predisposing elderly people to an increased risk of fracture (22). Bone strength seems to be determined by genetic as well as environmental factors. If genetic markers can indicate the risk of osteoporosis, they would be useful for its prevention and early, effective treatment. Several genes have been implicated as genetic determinants of osteoporosis (13, 16, 17). Recently, we identified a significantly stronger association between single nucleotide polymorphisms (SNPs) in the TNSALP gene (787T > C) (rs3200254) associated with bone mineral density (BMD) among 501 postmenopausal women (10). We genotyped two single nucleotide polymorphisms (787T > C [Tyr246His] and 876A > G [Pro275Pro]), which were shown to be in complete linkage disequilibrium. There was a significant difference in the BMD among haplotypes (P = 0.041), which was lowest among 787T/876A homozygotes, highest among 787T > C/876A > G homozygotes, and intermediate among heterozygotes. In subgroups divided by age, haplotypes were significantly associated with BMD in older ( > 74 years; P = 0.001), but not in younger (≦ 74 years; P = 0.964) postmenopausal women (10). These results indicate that the effect of haplotypes on BMD depends on age. Expression of the 787T > C TNSALP gene using COS-1 cells demonstrated that the protein translated from 787T > C exhibited ALP-specific activity similarly to that of 787T (10). Interestingly, the Km value for TNSALP in cells transfected with the 787T > C TNSALP gene decreased significantly compared to that of COS-1 cells bearing the 787T gene, reflecting the higher affinity. In the present study, we examined the biosyntheMolecular effects of TNSALP SNPs 215 Three-dimensional structure of human TNSALP. The modeled structures of human TNSALP were constructed using the program Discovery STUDIO, Modeling 1.1 (Accelrys Inc., CA, USA) based on human placental ALP (PLAP) (21). The amino acid substitutions in the three-dimensional modeled structures of PLAP and TNSALPs (787T, 787T > C) were investigated. Statistical analysis. Data are presented as means ± SE or means ± SD. Quantitative data were analyzed by Student’s t-test. Significance was considered at P < 0.05. Analyses were conducted using SPSS 13.0J (SPSS Inc., IL, USA). RESULTS AND DISCUSSION Metabolic labeling and immunoprecipitation To analyze the protein translated from the TNSALP gene 787T > C associated with BMD, we examined in vitro translation using the TNSALP cDNA expression plasmid (787T or 787T > C). Cells transiently expressing TNSALP (787T or 787T > C) were metabolically labeled with [S] methionine for 0.5 h and chased for up to 4 h. After 1 h of being chased, TNSALP (787T) and TNSALP (787T > C) similarly became an 80-kDa form (Fig. 1). Oda et al. reported that TNSALP migrates to the Golgi complex and becomes an 80-kDa form, as described previously (14, 15, 19). Therefore, the present results suggested that both proteins are similarly synthesized as a 66-kDa form, and migrate to the Golgi apparatus. The degradation of TNSALP (787T or 787T > C) was not observed, even in the 4-h chase (Fig. 1); this suggested that both proteins are stable. Expression of the investigated SNPs (787T or 787T > C) in TNSALP cDNA The stromal cell line ST2 cells did not express a significant amount of endogenous TNSALP. Previously, we reported that Pi starvation or BMP-2 induced the TNSALP activity and regulated its expression in ST2 cells (8, 9). Therefore, we examined the effects of amino acid substitutions on the catalytic properties using ST2 cells transiently expressing SNPs of the human TNSALP (787T or 787T > C) gene. As shown in Fig. 2, there was no significant difference in the levels of ALP-specific activity between 787T and 787T > C. When cotransfected with the same quantity of 787T and 787T > C plasmids (1 : 1), the cells exhibited a level of ALP-specific activity almost the same as that of 787T or 787T > C, and no interaction in ALP activi(27). Enzyme activity and protein assays. The mouse stromal cell line ST2 (Riken Cell Bank, Tsukuba, Japan) was cultured in MEM-α medium (Gibco, Invitrogen Corp.) containing 10% (v/v) fetal calf serum. ST2 cells were spread on 24-well pl


Introduction
Scientific communication and scholarly publishing are in transition.The age of printed publications as primary means to communicate research results is ending, being replaced by the era of electronic publishing (also known as e-publishing).This form of publishing has far-reaching consequences not only for how scientists distribute, access, process and digest information but also for how research itself is done and will be evaluated.
The advantages of electronic publishing are immediately evident: research results can be disseminated faster and more cheaply, can be distributed to a wider audience more fairly (it offers equity of access, including the lay public and scientists in developing countries) and authors have virtually no space restrictions, and can therefore include huge datasets or even multimedia data.It is obvious that information is key to research and knowledge production.The famous phrase coined by American sociologist Robert Merton -"Standing on the shoulders of giants" -actually refers to scientists using past work in advancing knowledge.If information is so crucial, certainly faster and cheaper dissemination of and access to electronic information should lead to better research.
However, not all scientists share the enthusiasm of having yet more information at their fingertips, in particular if this seemingly comes at a cost of quality.The problem of 'excessive publication' and information overload in immunology was already decried 20 years ago [1] and since then the number of immunology journals has almost tripled; in addition, scientists now have access to an unprecedented amount of information on the Internet.Unfortunately, more information does not always mean better information.For example, information on the Internet is often reported as being of poor relevance and validity [2,3].The recent outcry of many scientists, including The American Association of Immunologists (AAI), about having preprint servers for biomedicine (see Box 1) was partly driven by the fear of getting burdened by an avalanche of non-peer-reviewed electronic junk-science that is impossible to cope with.With this article I will gently oppose this view and argue that electronic publication in research actually refers to two different processes: firstly, sharing data and intermediate results for collaboration and discussion, where speed and relevance are more important than in-depth prepublication peer-review; secondly, communication to bring reasonably validated research results into practice.By making this distinction, the absurdity of the opposition to preprint servers, which contain non-peer-reviewed material, becomes clear.

What is electronic publication?
'Publication' literally means 'making public' and the word 'electronic' refers to information that is stored only in computers.Electronic publishing in the broadest sense can therefore mean many different things: I will give five examples.
The first example is papers that have already been published in print journals and that are in addition adapted into electronic form, published for example by electronic publishers such as HighWire Press at Stanford University (www.highwire.org).HighWire, which started in 1995 with the online production of the weekly Journal of Biological Chemistry, today offers more than 150,000 free full-text articles from more than 200 printed journals.Also in this category belongs electronically 'self-archived' material that has appeared elsewhere in print, for example authors of scholarly papers publishing their work on their homepages, or universities building up databases with theses and research reports.The second example is scientific papers published exclusively electronically (e.g. on the World Wide Web), either by the authors themselves (e.g. on their homepages, without peerreview) or by peer-reviewed electronic journals.The first biomedical journal that was published exclusively electronically was the Online Journal of Current Clinical Trials, which started in 1992.The third example is drafts of scientific papers submitted by authors and published in so-called preprint databases (also referred to as 'e-print servers'), such as Netprints ( [4]; see Box 1).The fourth example is publication of data and information in databases, for example nucleotide sequences in the EMBL/Genbank databases.The fifth example is that, in a broader sense, even the publication of meta-information -such as bibliographic information stored in databases such as Medline -may be referred to as electronic publication.
It should be noted that a grey area exists between what constitutes electronic publication and what doesn't, depending on what is considered 'public'.For example, if a researcher circulates a manuscript among a few colleagues via e-mail, not many people would actually consider this as 'electronic publication' whereas posting a manuscript to an electronic mailing list with hundreds of subscribers or publishing it on a preprint server or a website may already be considered electronic publication.This would result in certain journals following the so-called

The impact of preprint servers and electronic publishing on biomedical research Gunther Eysenbach
Ingelfinger rule to reject the article due to prior publication [5].But how 'public' does a document have to be to constitute 'publication'?The New England Journal of Medicine once made clear that "…posting a manuscript, including its figures and tables, on a host computer to which anyone on the Internet can gain access will consti-tute…publication.On the other hand, sending manuscripts by e-mail to a limited number of colleagues -a dozen or two, let us say -will not."[6].If 24 readers are not sufficient to create a 'public', how many readers are needed to constitute 'publication'?In fact, different journals have different policies on what they consider prior publication; for example Nature sees publication of sequences in electronic databases or draft manuscripts on preprint servers as part of the scientific communication process: "Genomics databases, like preprint servers and conferences, represent a form of intracommunity networking from which all researchers benefit.Nature does not count them as prior publications."[7].

Type-1 and type-2 electronic publications
Much confusion and misunderstandings arise if people speak about electronic publishing and actually mean different things.Whereas traditional publication was a much better-defined dichotomous event, with a clear mission of transporting research results to the scientific community and the public, publication in the electronic age is much more a continuum [8] reflecting, and occurring during, the entire research process from hypotheses formulation to data gathering, raw data interpretation and the presentation and discussion of the final data.The more 'collaborative' research has to be, the earlier in this process electronic communication and 'publication' should occur.Electronic publishing in a broader sense includes the whole spectrum of electronic communication during the research process -for example, generating and sharing protocols and electronic draft data or draft manuscripts with research colleagues -whereas electronic publishing in a narrower sense refers only to the final, peer-reviewed release of data as a culmination of a research process.
It is important to discriminate between these two very different concepts of electronic publication.In the following I will refer to the former (electronic data released as part of the scientific collaboratory working process) as type-1 electronic publication and to the latter (carefully peerreviewed electronic publication as a preliminary endpoint of a project) as type-2 electronic publication.
Type-1 electronic publications are characterized by opening work-in-progress to colleagues, thereby improving collaboration and quality.Typically 'published' information are draft data that need to be shared quickly among researchers, perhaps on a global scale, or preliminary

Box 1 Preprint servers
Early experiments of distributing preprints and other 'type-1' communications among scientists in written form were conducted in 1961 by the US National Institutes of Health (NIH) and called 'Information Exchange Groups'.In the pre-Internet era scientists received photocopied material, which was a very costly process and which led to an end to the experiment in 1966 [19].
Electronic preprint servers evolved in the field of physics from August 1991 onwards and are now in many research areas an established medium to communicate non-peer-reviewed results of ongoing research among researchers.Preprint servers are actually Internet-accessible databases; they allow scientists to deposit electronic draft articles in order to make them accessible to a wider academic audience, before they actually submit them to a peer-reviewed journal.Strictly speaking, 'preprint' is a grossly misleading term because it suggests that papers published on these servers will eventually be 'printed' , which may not necessarily be the case: firstly, it is not certain whether papers published on preprint servers will ever be submitted or accepted for publication at all; and secondly, if they are accepted by a peer-reviewed journal, they may well end up in an electronic journal and not necessarily in a printed journal.The term 'e-print server' (which is somewhat oxymoronic in combining the terms 'electronic' and 'print') may be even more confusing.Thus, when using the term 'preprint' we actually mean 'pre-peer-review' or 'pre-submission' documents.
The preprint server in the field of physics -formerly known as the 'xxx preprint archive' (xxx.lanl.gov,now known as ArXiv.org)-today serves 25 research disciplines, such as high-energy physics, economics, and atmospheric and oceanic sciences.
On 22 April 2000, The NIH director, Harold Varmus, and colleagues David Lipman (Director of the National Center for Biotechnology Information) and Pat Brown (a geneticist at Stanford University in Palo Alto) circulated a proposal for the first preprint server in the field of biomedicine; the server was first named 'E-Biomed', later 'E-biosci' and is now known as 'PubMed Central' (http://pubmedcentral.nih.gov)[20]."Taxpayers have paid for research already, so NIH should make the results widely available…" was one of the arguments for the Varmus proposal to establish PubMed Central [21], which originally was not only meant to become a electronic repository for already published research but also was supposed to contain a preprint section that allowed researchers to submit papers directly without peer-review.This latter part of the proposal soon came under severe fire.The New England Journal of Medicine, which has earlier already argued that "…publishing preprints electronically sidesteps peer-review and increases the risk that the data and interpretations of a study will be biased or even wrong." [6], published an editorial pointing out that "The best way to protect the public interest is through the existing system of carefully monitored peer-review, revision, and editorial commentary in journals." [22].
As a result of the fierce criticism from scientific publishers, the NIH later dropped the idea of an electronic preprint server containing unreviewed material [23] and currently PubMed Central seems to have become an electronic platform to distribute full-text papers that have already been published in traditional journals or that have gone through peer-review by an editorial board (in other words, a platform primarily for type-2 communications).Meanwhile, the European Molecular Biology Organization (EMBO) also decided to create a free website, named E-Biosci, as a portal site for life-science papers; this is the European counterpart to PubMed Central [24].
However, the idea of a preprint server to serve type-1 communication has already been taken up by the several commercial publishers: for example, the British Medical Journal Publishing Group together with the Stanford libraries launched www.netprints.com as a preprint server for the entire field of medicine [4,25].
results that could benefit from the input of a broader research community.Genome databases containing draft nucleotide sequences are a typical example but so are preprint servers.Type-1 electronic publications have a similar validity to papers presented at conferences: they have not gone through a rigorous peer-review process but are primarily discussed during or after 'publication'.The very process of type-1 electronic publication is aimed at providing input from a broader research community.The emphasis of type-1 communications is not on validity (the reader is aware that he is dealing with draft data) but on openness and speed.Researchers look at these electronic publications because the results, despite being tentative, may be relevant to their own work.Researchers are expected to do their own 'downstream-filtering' of relevant information, which in the electronic world can be facilitated by providing meta-information [2].
Type-2 electronic publications are different.Their aim is to bring reasonably valid research results into practice.The results have important implications and they are expected to be acted upon on a wider scale.They may lead to changes in clinical practice or to policy changes.They are the preliminary endpoint of a long process of careful research, discussion and rigorous peer-review.The publication of a clinical trial in the New England of Medicine is a good example.Here the emphasis clearly lies on validity; 'upstream-filtering' in the form of peer-review prior to wide distribution is important.
Type-1 and type-2 communications are, in the electronic world, more difficult to discriminate from each other than in the traditional publishing world, where 'publication' was inevitably linked with the notion of peer-review and quality control and therefore immediately recognizable as type-2 communication.Unlike traditional publication, in type-1 and type-2 electronic publication the two processes of improving the quality and making the paper physically available are two distinct processes [9].They may even occur in the opposite order as compared with traditional publishing -an ongoing peer-review process after publication is possible (e.g. by HighWire's 'rapid responses' or 'post publication peer-review [P3R]', as the journal Pediatrics calls it).
The confusion that arises if people fail to acknowledge that type-1 and type-2 communications are two different things can be best illustrated by the many responses to the PubMed Central proposal of having a preprint server for biomedicine (see Box 1).Among others, representatives of the American Association of Immunologists felt that the "…proposal compromises the cornerstone of scientific method: peer-review.The process described in your proposal…does not ensure a rigorous peer-review process.Without this we compromise our excellence (at best) and (at worst), pose potential harm to the scientific community as well as the public at large.Furthermore, scientists depend on the current peer-review process to give their work legitimacy and guidance; they do not want to be held to lesser standards."[10].
The concern here was that by having type-1 and type-2 communications on the same server, the non-peerreviewed section would 'contaminate' and compromise the quality of type-2 communications.Other opponents of the proposal felt that readers could have trouble in distinguishing the different sections.Proceedings of the National Academy of Sciences of the USA felt that "…making non-peer-reviewed as well as peer-reviewed material available will confuse both scientists and the public…" [11].However, this perhaps belittles the ability of scientists to recognize different levels of evidence and to be able to interpret labels that could make clear that certain material is non-peer-reviewed content, as used in Netprints -after all, "…this is the age of transparency rather than paternalism…" [8] as Richard Smith, editor of the British Medical Journal, put it.

The benefits and problems of type-1 'open' electronic publication
One example of the benefits of open communication and data sharing comes from the 'open-source software' industry.This comprises computer programs, and developers freely distribute the source code and allow usage and modification.The Open Source Initiative explained the concept as follows: "The basic idea behind open source is very simple.When programmers on the Internet can read, redistribute, and modify the source for a piece of software, it evolves.People improve it, people adapt it, people fix bugs.And this can happen at a speed that, if one is used to the slow pace of conventional software development, seems astonishing…We in the open-source community have learned that this rapid evolutionary process produces better software than the traditional closed model, in which only a very few programmers can see source and everybody else must blindly use an opaque block of bits."(http://www.opensource.org/).Replace 'software' with 'research' and 'programmers' with 'scientists', and you have a perfect justification for type-1 open-source publishing.It is also noteworthy that the initiative says that "The foundation of the business case for open-source is high reliability.Open-source software is peer-reviewed software; it is more reliable than closed, proprietary software.Mature open-source code is as bulletproof as software ever gets.".
Although, during the development process, open-source code may appear immature, preliminary, non-peerreviewed and of lower quality than commercially available software, the software industry has learned that the endproduct of open-source development is of superior quality.What is true for the software industry has strong parallels to the area of research: perhaps its strongest analogy in the field of genomics [12], where it is (according to the socalled Bermuda agreement) common practice for researchers to place sequence data on public and freely accessible databases as sequences are generated (nonpeer-reviewed and in a draft status).The analogy may also be extended to preprint servers, which allow research protocols, draft papers and datasets to be published and reviewed by others, who could give valuable input.
In certain areas, such as in genomics research, type-1 electronic communication is a necessity to foster international collaboration.In other areas, for example clinical research, electronic publication can also help reviewers who attempt to synthesize research in an unbiased manner.A current problem for authors of reviews on the effectiveness of a clinical intervention is that the literature may be biased in favor of positive or promising results, which are more often published in paper journals than negative results (this is known as publication bias).This may affect the validity of systematic reviews [13].Electronic registers of clinical trials (another kind of type-1 electronic publishing), where investigators publish their research protocols from the early stages onwards, can later help to identify negative trials that remained unpublished [14].
It is likely that, as in clinical research, many results in experimental research are only published if they are desired or significant.In experimental research, preprint servers could play a similar role as prospective trial registers in clinical research: scientists can deposit protocols of ongoing experiments and briefly report findings electronically that otherwise would not deserve publication, thereby providing a perhaps more genuine picture of reality.
Despite these considerations, it must however be acknowledged that openness also brings at least three problems concerning intellectual property issues: firstly, debates over priority, authorship and credit for analyzing draft data in depth that have been made entirely open may arise [7]; secondly, the danger of plagiarism from non-peerreviewed electronic material that has been made public [15]; and, thirdly, the problem that European patent laws (contrary to US laws) do not allow the patenting of data that have been published [16].Therefore, not all material is suitable for preprint servers.

Conclusions and outlook
Totally new concepts of 'publishing' and distributing data will evolve in the near future.Type-1 electronic publishing may be become a subtle process that will have nothing in common with what we traditionally know as publication.One example is software using so-called 'Napster technology' that allows searching for certain data across the hard disks of all scientists who are willing to share their data.This kind of software is already envisaged to help the annotating of genome sequences in a collaborative way.[17] The simple act of a scientist marking one of his files as publicly accessible may already constitute publication.These developments also challenge the way that research currently is being evaluated.In their criticism of the PubMed Central proposal, The American Association of Immunologists wrote that presently "…scientists depend on the hierarchy of journals to help them select the most important studies in the plethora of information available to them.It is unclear how a single information source would assist this sorting process."[10].Clearly, traditional methods to assess the value of research -such as journal impact factors -will become redundant [18]; however, new methods will evolve.The value of a manuscript will become more important than the impact factor of the journal in which it is published.Electronic publishing will provide alternative models, for example a 'paper auction' model: researchers could submit type-1 electronic papers to preprint servers for discussion and peer-review, and journal editors and publishers would pick and bid for the best papers they want to see as 'type-2 papers' in their journal.The best journals would be able to pay the highest prices for the best papers and the number of bidders or the sum that was bid for each paper would determine its value.
As the number of projects that all share the common goal -to improve electronic scholarly communication -is increasing, co-operation and interoperability between these developments are becoming key challenges.Although Internet technology provides the basic protocols to link different services physically, higher-level standards are needed to ensure interoperability.The Open Archives initiative (www.openarchives.org) has recently taken a first step in proposing a convention that provides a technical and organizational framework to support basic interoperability among e-print archives.
The costs shift away from publishing and distributing information, and towards finding and managing relevant and valid information.Accessibility and connectivity of information need to be improved: in type-2 publishing, data are filtered upstream (by means of peer-review) whereas in type-1 publishing, scientists need to be able to select and filter relevant information downstream, which requires labeling with computer-readable meta-information [2].
In an ideal future, researchers should be able to browse through a global knowledgebase in order to search across different literature databases, full-text archives and digital libraries, and to navigate seamlessly from one publisher's server to another and from database producers and preprint servers.