March 1998 Volume 5 Number 3 p165
In a series of e-mail messages Philip Campbell, Floyd Bloom and Benjamin Lewin, the editors of Nature, Science and Cell, have been discussing whether they should no longer allow authors of papers describing three-dimensional structures of biological macromolecules to put the coordinates on a post-publication one year hold in the Brookhaven Protein Data Bank (PDB). Independently, a similar debate is underway at the National Institutes of Health (NIH) in the USA, involving Harold Varmus, the Director of the NIH, Marvin Cassman, the Director of the National Institute of General Medical Sciences (NIGMS), and Howard Schachman, the NIH ombudsman.
The International Union of Crystallographers (IUCr) is also taking close look at its guidelines for deposition, now almost a decade old. Ted Baker, the President of the IUCr has "asked Wolfram Saenger, the current Chairperson of the IUCr Commission on Biological Macromolecules, to initiate a discussion on whether our guidelines are still appropriate or whether they should be revised". Baker adds that the consultation process should be complete well within six months, assuaging fears that the process could take a year or more1. Even so, it is clear that many of the journals that publish structures, as well as the NIH, are ready to move quickly, in the absence of new guidelines from the IUCr, should these be slow in forthcoming.
Those at the NIH are champing at the bit. "How can you rationalize public support for research if the results of that research are not then made public on publication? I think there is widespread support both within the NIH and in the community for no longer allowing the one year hold," says Cassman. Schachman has been catching some of the fall-out from the NIH's policy, clearly, when he says, "I have heard a lot of complaints about failure to deposit, as well as coordinates being on one year hold". Cassman is determined that things should change: "This time the NIH will go ahead without the IUCr. A lot has happened since the last guidelines were proposed. The utility of three-dimensional structures has become much broader than it was a decade ago: many people find coordinates necessary and useful for their research, they provide an essential part of understanding and utilizing the data".
But Peter Wright, the editor-in-chief of the Journal of Molecular Biology (JMB), and Wayne Hendrickson, the joint editor of Structure1, are keen for the IUCr to take the lead. Wright says that JMB "will go along with what is decreed by the IUCr". Indeed, most of the other journals presently abide by the IUCr recommendations, a situation not all are comfortable with. Nick Cozzarelli, the editor-in-chief of the Proceedings of the National Academy of Science of the USA (PNAS) laments that "[structural biologists] held us hostage, and the journals were pusillanimous in the face of this".
Surprisingly, Cell has the strictest policy of all, at least in theory. Lewin clarifies: "There is nothing in Cell's policy that says you can have a one year hold." Does Lewin then chase researchers who publish in Cell to release their coordinates if they put them on hold? "No, it's not quite like that -- no one has ever complained to me about coordinates of a structure published in Cell being on one year hold. This is what the crystallography community has tacitly accepted among themselves, for whatever reason. But, if you come to Cell and say you want a one-year hold, we would say that is unacceptability long. If, on the other hand, you come along and say that there is a paper in Cell and you can't get access to the [coordinates], I will take that issue up with the author. There are so far no cases in which the authors have remained resistant."
None of the journals or the NIH wishes to move ahead without consulting structural biologists in some manner. "My big concern," says Wright, "is that the community has some say. For most of them [the change in policy] would not matter (they may complain about it but it is not going to affect their science directly) but there is a significant population out there, working in or with industry, who will be directly affected." Philip Szuromi, one of the senior editors at Science, wants a near unanimous vote in favor before any policy change is made: "What matters is that we get 98% of the crystallographers to sign on. If they don't sign on, then [by requiring immediate release of coordinates] we may be creating an artificial problem". Cassman is more pragmatic: "We are trying to get some kind of consensus within the NIH and in the community: it will not be 100%, I am not going to pretend that". Cozzarelli indicates that PNAS will rely on its editorial board for a decision on its policy regarding release of coordinates. "There are four structural biologists on the board: Roger Kornberg, Tom Pollard, David Davies and Don Wiley. All four say that upon publication the coordinates should be available, not [just] deposited, available. So, given that the four members of the board who are experts are unanimous, I would be stunned of the rest did not follow."
It seems unlikely that PNAS or any of the other journals will make the move unilaterally, for this would place them at a potential competitive disadvantage relative to the other journals in attracting papers. Lewin is particularly sensitive to the issue. "I think dropping the one year hold is a perfectly sensible idea and I am happy to lend some force to it. Our main concern about tightening our policy is that we put ourselves at a competitive disadvantage. We have moderately bitter memories of the days of cloning when Cell demanded that clones were handed out and Nature, under [John] Maddox, said we don't care, do what you like. We lost a lot of papers and many labs said to us quite overtly that they were going to publish in Nature because they didn't have to provide the clones. We stuck to our policy because we thought it was right and, frankly, I thought Nature's behavior was disgraceful. I am not terribly keen to put myself in the same position again with regard to coordinates."
Campbell, who took over from Maddox two years ago, is adamant that Lewin's fears are misplaced: "[It has been suggested] that Nature is laissez faire for competitive reasons. Not so while I have been editor: when we changed Nature's policies on deposition over a year ago, we explicitly kept in line with community custom in allowing the one year delay. I am keen to explore changes to our policy, subject to ensuring we don't give other journals an advantage."
Lewin notes that "the critical thing is that the journals all take the same attitude to an offense. The journals will have to agree ahead of time what they will do in cases where authors refuse to provide all the information on acceptance of a paper." This is where there may be the biggest philosophical differences between the journals. While JMB, Nature Structural Biology and Nature, for example, insist upon PDB accession numbers appearing in papers prior to publication, Cell and Science do not think this is the right approach to the problem. "Cell has only a limited amount of time and resources and one of the things I really don't want is to start chasing people for things like accession numbers. Nor am I going to hold up a paper in the publication schedule because the authors cannot provide the accession number in a timely manner for technical reasons from time to time." Szuromi is more explicit about potential sources of delay: "In cases where we publish papers rapidly, we are way ahead of the PDB granting an accession number."
Joel Sussman, head of the PDB, is quick to counter that "for the last year and a half, when you deposit your data in the PDB, there is a procedure, Autodep, that automatically e-mails you the accession number within 45 minutes of deposition. For European depositors going through the EBI it may take up to 24 hours. Autodep is available on Saturdays and Sundays and even Christmas Day. Following this, the PDB staff reviews the data carefully before release, often in consultation with the authors; this takes on the order of 100 days"2.
Cassman and the NIH want the journals to take a vigorous line when enforcing the present IUCr guidelines. "The journals have to say that they will not accept a paper unless it has a deposition number. Even now, I think there are a number of journals that say they do it but don't really implement it very vigorously." While the discussions continue about how Nature, Science and Cell could collaborate in changing their requirements for deposition and release of coordinates, it is clear that there are many important details that have to be resolved before they can present a united front.
The most important issue remains the response from the community itself. It is clear that the IUCr has the potential to play an important role in guiding the formulation of a new set of guidelines for deposition and release of structure coordinates. It is also clear that a number of people involved with the journals and funding agencies are sufficiently uncomfortable with the present policies to move ahead on their own. The only question then is whether the IUCr, or a similar international body, will be able to act in a prompt enough manner to influence the outcome of the debate.