correspondence Unrestricted free access works and must continue Bioinformatics researchers shouldn’t need coercion to act responsibly and collegially.

Sir — The policy on release of unpublished for all to use without restriction”. This goals of the scientists at the genome data from large genome centres has statement, which notably uses the word centres, who should be consulted as generated considerable discussion and “should” rather than “must”, is qualified part of any large-scale analysis of some confusion, as your Editorial by a lengthy discussion of conditions, unpublished genome data, and included “Sacrifice for the greater good?” makes including a reference to the sequence as collaborators where appropriate. It is plain (Nature 421, 875; 2003). In our view, producers’ interest in “the first peer- a serious problem that these invaluable data sets from large, centralized, expensive reviewed published analysis of the results centres feel compelled to coerce such genome data-collection projects should be of the sequencing project”. This reflects simple scientific courtesy from our freely available to the entire scientific the real concern of the genome centres community. We encourage our peers community, immediately and with no that prepublication data access may in bioinformatics to act responsibly, restrictions or conditions. put the scientists there at a competitive cooperatively and collegially, to help to Our position is that pre-publication disadvantage. We understand these assure open, unrestricted, immediate release of large genome data sets is a special concerns, but we believe that the release from large community-driven case, and not a principle that should be qualifying discussion cannot coexist with data-collection projects. applied “throughout the world of biology”, the principle of “without restriction”. We Steven Salzberg*, Ewan Birney†, as was asserted in your editorial. Large propose that these qualifications are Sean Eddy‡, Owen White* genome sequencing has become an simply dropped to avoid confusion among *The Institute for Genomic Research, 9712 Medical expensive, factory-style operation, in data consumers, journals and journal Center Drive, Rockville, Maryland 20850, USA which economies of scale can only be reviewers. The Human Genome Project [email protected] realized by very large centres. Large data has been a spectacularly successful †European Bioinformatics Institute, production centres, established and demonstration that the “Bermuda rules” Wellcome Trust Genome Campus, Hinxton, supported by the scientific community, of free access without restriction do work, Cambridge CB10 1SD, UK represent a different model of science from for everyone. ‡Howard Hughes Medical Institute and traditional investigator-initiated projects. As bioinformaticians, we have an Department of Genetics, Washington University We argue that they need to operate under important role in this process. We reaffirm School of Medicine, 4566 Scott Avenue, different rules. our own commitment to respecting the St Louis, Missouri 63110, USA The broader scientific community supports the highly centralized model represented by the US large-scale genome Watching an IT icon insistent photographer that the slide rule centres (funded via the National Institutes was irrelevant.) of Health and the Department of Energy) slide into history At Newcastle University we have been on the condition that everyone in this teaching biology students the use of community gets equal access to the data. Sir — In the famous photographs of computers for nearly three decades, If this is the case, everyone wins: large data James Watson and Francis Crick with beginning each year with a brief account of sets are generated at the lowest cost and their DNA model (reproduced in Nature the staggering advances in computational greatest speed, and scientific work 421, 15 & 417; 2003) a small icon of power since the middle of the twentieth progresses on multiple fronts without information technology can be seen in century. Each year, we show them an delay. In contrast, if genome centres Crick’s right hand — a slide rule. (One example of what Crick held. It seems restrict their data and get preferential wonders, incidentally, why this was fitting, on the fiftieth anniversary of the access to it, then some members of the shown in the photograph instead of the reconstruction of DNA, and at the community will no longer support “measuring stick” which, together with advent of the rise of computational monopolistic funding models (in which a plumb line, they used “to obtain biology (C. Surridge Nature 420, 205; large centres sequence one genome after the relative positions of all atoms”, as 2002), to report that this year, for the first another without peer review of each described by Watson in The Double Helix time, none of our first-year students project). Instead, they will demand the — anecdote has it that Crick told the recognizes a slide rule. right to compete with these empires, W. John Cram especially for the most scientifically School of Biology, University of Newcastle upon desirable genomes. Other scientists, Tyne, Newcastle upon Tyne NE1 7RU, UK especially bioinformaticians, will seek to relocate to the centres to gain the advantage of early data access. Data restrictions will therefore promote Secrecy is increasing in factionalization where we should step with competition BROWN/SPL A. BARRINGTON be seeking efficiencies of scale, and centralization where we should be Sir — Various authors and reports have promoting diversity. recently claimed (for example, refs 1–3) We agree with your editorial that the that the increasing commercialization of proposed new policy, recently released for academic science has led to an increase comment by the US genome centres (see in secrecy. However, our comparison of www.genome.gov/page.cfm?pageID= two surveys of experimental biologists, 10506537), is ambiguous. It states that Ruled out: unlike Francis Crick, it seems that the mathematicians and physicists, conducted genome sequence data “should be available slide rule is slowly being forgotten. about 30 years apart, suggests a more

NATURE | VOL 422| 24 APRIL 2003 | www.nature.com/nature © 2003 Nature Publishing Group 801 correspondence complicated and interesting picture. Recent increases in US government related basic science research, as is done As feared, secrecy (measured as funding for science, if they are sustained, in China3; and training researchers in unwillingness to discuss ongoing research may help to lower the intensity of state-of-the-art skills, along the lines of with those outside the research group) competition, as well as the dependence initiatives by the European Molecular has increased. on industry funding, and thereby reduce Biology Laboratory. In 1966 (ref. 4), 50% of 1,042 secrecy. Furthermore, although we need to All the stakeholders — in the fields respondents reported feeling safe in be wary of the strings attached to industry of policy, administration, science and talking with all others about their current funding, university– industry collaborative industry — have to address this problem research, but by 1998, when we surveyed research should be encouraged. and give a directional nudge to research 202 scientists from the same three fields We thank Lowell Hargens for providing initiatives. (details of methods and results available field-level data from ref. 4. Nature has covered some of these issues from J.W.), the equivalent number was John P. Walsh*†, Wei Hong† recently, for example the low impact 26%. Experimental biologists have become *Research Center for Advanced Economic factors of Indian biological-science particularly secretive, with only 14% Engineering, University of Tokyo, 4-6-1 Komaba, journals4; the urgent need to explore the willing to talk openly about their current Meguro-ku, Tokyo 153-8904, Japan best scientific options for sustainable research in 1998. Secrecy is strongly [email protected]. development by regional centres of the predicted by scientific competition †Department of Sociology, University of Illinois, International Council of Science5; and (measured as concern over having one’s Chicago, Illinois 60607-7140 USA the effect of post-genomics research on 6 research results anticipated). The effects 1. Blumenthal, D. et al. J. Am. Med. Assoc. 277, traditional methods of food production . of commercial activity, on the other hand, 1224–1228 (1997). Informal articles on topics such as these 2. Cook-Deegan, R. M. & McCormack, S. J. Science 293, are quite mixed. Patenting has no effect; 217 (2001). serve as a compass for framing policies industry funding is associated with greater 3. Campbell, E. G. et al. J. Am. Med. Assoc. 287, 473–480 (2002). and making course corrections during secrecy; but having industry collaborators 4. Hagstrom, W. O. Am. Sociol. Rev. 39, 1–18 (1974). implementation. is associated with less secrecy. 5. Cohen, W. M., Nelson, R. R. & Walsh, J. P. Management Sci. R. M. Ranganath 48, 1–23 (2002). These university–industry collabo- Department of Botany, Bangalore University, rations can be viewed as part of a strategy Jnanabharathi Campus, to share findings and expertise with the Bangalore 560056, India wider scientific and technical community. 1. Parthasarathi, A. Nature 422, 17–18 (2003). For companies, timeliness and Pre-genomics training 2. Palsson, B. Nature Biotechnol. 20, 649–650 (2002). customization of information are often 3. Cyranoski, D. Nature 410, 10–12 (2001). hinders Indian biotech 4. Vohora, S. B. & Vohora, D. Nature 412, 583 (2001). more important than exclusivity, so they 5. Clarke, T. Nature 420, 733 (2002). are willing to tolerate, even encourage, Sir — Ashok Parthasarathi in his 6. Knight, J. Nature 421, 568–570 (2003). their academic collaborators’ participation Commentary “India: a champion of new in the shared conversation of a scientific technologies” (ref. 1) rightly affirms India’s field, thereby giving the company access bid to play a leading part in global to the whole community’s expertise. technology developments. India’s Should tobacco ban rule In contrast to these collaborations, technological independence has come industry funding alone is often associated mainly in the physical sciences, such as out governments too? with a university laboratory acting as a space research, telecommunications, Sir — I was interested to read in your News subcontractor to a company’s R&D software, defence systems, energy and story “Academics fume as university project, and may produce associated supercomputing. refuses to reject tobacco dollars” (Nature secretive behaviour. Although Parthasarathi lists biotech- 422, 361; 2003) a scientist at the University Thus, there is reason to believe that nology as a key ‘champion technology’, a of California, San Francisco, quoted as secrecy has increased among academic tremendous boost is needed for it to make saying it is immoral to accept research scientists, but that the focus on a global impact. funds from tobacco companies because commercialization as the cause may be India shares this predicament with “it is not appropriate to take money from misplaced. Although commercial activity other developing countries that have an industry that kills 5 million people may reduce formal activities such as a vast research force but the training worldwide and constantly lies”. publication or sharing of materials, it and infrastructure of pre-genomics I guess that means that no research may have fewer negative effects on days, and little experience of the research funds should be accepted from the US informal communication among skills required for post-genomic biology. Department of Defense, nor from its researchers. As this informal communi- With the genome sequence data for corporate headquarters — the US cation is significant in transferring several organisms, including humans, government. Other national governments information to companies5 and is at now available, the rules of knowledge- would presumably also be banned from least as important as publication for based commercial ventures have changed supporting research under these new distributing information among scientists, and an era of ‘omics’ — genomics, politically correct guidelines. this is encouraging news. proteomics and so on — has emerged2, Richard S. Jope Although it is right to raise concerns with functional genomics as the new Department of Psychiatry, about the negative effects of publication ‘mantra’. Sparks Center 1057, 1720 7th Avenue South, restrictions associated with industry In order to become a champion of University of Alabama at Birmingham, funding, we should not conclude that biotechnology, India needs a paradigm Birmingham, Alabama 35294-0017, USA university–industry linkages per se shift in the organization of research and produce unhealthy levels of secretiveness training, priming research institutions correspondence among academic scientists. Instead, it and universities to change gear and meet Contributions to Correspondence may be submitted may be better to focus on alleviating global challenges. Two aspects need to [email protected]. They should be no longer some of the negative consequences of urgent attention: establishing functional- than 500 words, and ideally shorter. Published scientific competition. genomics centres for biotechnology and contributions are edited.

802 © 2003 Nature Publishing Group NATURE | VOL 422 | 24 APRIL 2003 | www.nature.com/nature