Three papers caught up in journal probe of review process.
A dispute between the editorial board of the Proceedings of the National Academy of Sciences (PNAS) and an academy member has put the fate of three studies in question. In the wake of rows over a controversial paper published by the journal online in August — but not in print — two additional papers linked to the same academy member are now in limbo.
Last month, PNAS editor-in-chief Randy Schekman wrote to academy member Lynn Margulis, a cell biologist at the University of Massachusetts, Amherst, asking for "a satisfactory explanation for [her] apparent selective communication of reviews" for a paper she ushered through the peer-review process. Schekman made the demand after a report in Scientific American cited Margulis as saying that she obtained "6 or 7" reviews before netting "2 or 3" favourable ones that recommended publication.
The paper in question, by Donald Williamson, a retired zoologist at the University of Liverpool, UK, claims that the transition of caterpillars into butterflies can be explained by ancient butterflies inadvertently mating with velvet worms1 . This controversial idea is supported by Margulis, who is a strong proponent of the hypothesis that new species form by symbiotic mergers between unrelated organisms. She denies any wrongdoing and stands by the work.
But Williamson's claims met with scepticism from many scientists after the paper was published online. "If you know the literature on insect metamorphosis and insect development, you would know right away that this is absolutely ridiculous," says Fred Nijhout, an insect developmental biologist at Duke University in Durham, North Carolina.
In last month's letter to Margulis, Schekman says that Williamson's study would not be printed, and that another paper, co-authored by Margulis and already accepted for publication, would not move forward until his concerns were addressed. That paper, led by Øystein Brorson of the Vestfold Hospital in Tønsberg, Norway, and co-authored by Margulis, describes a novel antibiotic treatment for Lyme disease, which is caused by primitive, spiral-shaped bacteria called spirochaetes — one of Margulis's areas of expertise.
Brorson says that he submitted page-proof corrections on 31 August and that the paper was slated for publication on 7 September. The new treatment, he claims, is 800 times more effective than doxycycline, the existing drug of choice against the disease. "It is very important to get this paper published," he wrote in an e-mail to Nature.
Of course I'm not withdrawing any of these papers at all. Lynn Margulis , University of Massachusetts
Now another paper put forward by Margulis is also being challenged, Nature has learnt. The paper, by John Hall, a computational biologist based in New York City who is an adjunct professor in the same department as Margulis, argues that genes from spirochaetes contributed to the genomes of advanced organisms, further supporting the theory of trans-species mergers.
The paper was accepted by three anonymous reviewers, but was questioned by a member of the academy's board who found fault with Hall's method of comparing gene sequences. In a letter on 4 September, Schekman urged Margulis to withdraw Hall's paper.
"Of course I'm not withdrawing any of these papers at all," Margulis says.
Schekman declined to comment, citing the confidentiality of the review process. "We are working with Dr Margulis and our conversations are ongoing," says PNAS spokesman Jonathan Lifland. "We don't want to respond to any questions or complaints she has through the media."
Like Williamson's paper, Hall's study was 'communicated' to PNAS by Margulis through a system called 'Track I' that allows academy members to bring papers written by non-members to the journal's attention and then choose the reviewers. Last month, the journal announced that it would eliminate this option with effect from July 2010 — a decision not driven by Margulis's submissions, says Lifland. "It was bad timing."
Meanwhile, Margulis has replied to Schekman's complaints. In a 5 October letter, obtained by Nature, Margulis details the eight people she asked to review Williamson's manuscript. Three scientists submitted formal reviews and three researchers declined to evaluate the paper — two because they were unavailable, one because he felt the topic was outside his expertise. Two amateur-naturalists also offered their comments, although Margulis didn't originally include these critiques because she felt they would be disqualified from the PNAS review process owing to the reviewers' lack of formal credentials.
Margulis says she regrets the omissions, but stands by her decision to ask non-academics to share their views. "My modus operandi is to ask competent people, whether or not they have a PhD," she says.
Hall is preparing his own response to Schekman regarding the board member's critiques, which he says are misguided. "I don't think this guy has done his homework," Hall says. "I'm still very hopeful that [the paper] will be published." The journal's editorial board evaluates all PNAS submissions before final acceptance, regardless of the submission route.
"Extraordinary claims require extraordinary evidence," says Martin McMenamin, a palaeontologist–geologist at Mount Holyoke College in South Hadley, Massachusetts, who reviewed Williamson's paper and recommended publication. "But I'm willing to lower that bar," he says, because evolutionary biologists are an "entrenched group" who can be reluctant to "consider alternate ideas".
"We will win one way or another because this is science," Margulis wrote in an e-mail. "I followed all the rules and submitted more reviews than I needed, and if they definitively reject these papers I will make it very clear to the reading public (because they make it clear in their anonymous letters) that, as usual, they don't like my ideas."
Williamson, D. I. Proc. Natl Acad. Sci. USA doi: 10.1073/pnas.0908357106 (2009).
About this article
Cite this article
Dolgin, E. Row at US journal widens. Nature (2009). https://doi.org/10.1038/news.2009.985