Retracted retraction leaves Genomics paper intact — but authors wonder if anyone will know

Last March, the journal Genomics retracted a paper, “Discovery of transcriptional regulators and signaling pathways in the developing pituitary gland by bioinformatic and genomic approaches,” for reasons that don’t really fit into a tight lede sentence. Let’s just say that at times the problems involved both questions of authorship and the validity of the research. More on all that in a moment.

Meanwhile, things change. Now the journal, an Elsevier title, is un-retracting (that can’t be a real word, can it?) the retraction. You’d think that would please the authors, and it does to an extent. But they also wonder, legitimately, whether the original retraction will refuse to relinquish its grip on the resurrected article and consign it to database oblivion.

First, some background. The paper in question — which has been cited six times, according to Thomson Scientific’s Web of Knowledge — comes from the lab of Sally Camper, head of the human genetics program at the University of Michigan. The first author had been, but is no longer, Nicola Solomon, an Australian scientist who came to Camper as a post-doc in the mid-aughts.

Solomon might not be too keen on federal officials in her adopted country, because she eventually found herself the subject of an investigation by the Office of Research Integrity (ORI). Earlier this month, ORI released its final determination, complete with its talk of multiyear sanctions (two, in this case) on various aspects of conducting federally-funded research.

Here’s a taste of what the inquiry found:

[Solomon] did not perform DNA sequencing on 202 cDNA clones of homeobox genes to confirm their identity and integrity. Through multiple revision of the manuscript, the Respondent did not discuss this with the corresponding author or question and correct the corresponding author’s addition of text indicating that the clones had been fully sequenced and were full length or longer (as indicated in Table 3) when compared to NCBI Mus musculus Unigene. This text supported the use of the Cap-Trapper technique to produce full length clones for the discovery of new genes without polymerase chain reaction (PCR).

Both the Respondent and the U.S. Public Health Service (PHS) are desirous of concluding this matter without further expenditure of time and other resources and have entered into a Voluntary Settlement Agreement to resolve this matter. This settlement is not an admission of liability on the part of the Respondent.

What does all that mean? Two things. Solomon said she did some genetics work but didn’t. And, it sounds as though the ORI decided they’d pushed the case as far as they could afford to, and got the scientific misconduct equivalent of a plea bargain without any admission of guilt. ORI typically sanctions researchers for three years, although we’ve seen other cases in which they’ve given out two-year sentences.

Camper said she first recognized a possible problem with Solomon’s research during the process of making changes that reviewers for Genomics had requested on the manuscript. By that time, Solomon had already left Camper’s lab for the University of Southern California — she’s no longer there, either — so Camper asked another researcher to help satisfy the reviewers. That scientist — Michelle Brinkmeier, the second author manuscript, who had initiated the project before Solomon joined the lab — could not replicate Solomon’s work, which was supposed to have been sequencing clones that had been provided by a lab in Japan.

Upon resubmitting the corrected manuscript, Camper informed Genomics that she would be removing Solomon as a co-author — putting her into the acknowledgments section instead — and adding Brinkmeier in her place. That change — and Solomon’s hostile reaction to it — explains the rather confusing retraction notice that appeared in March of this year, nearly two years after the paper was published:

This article has been retracted at the request of the Editor-in-Chief. A co-author, who was indentified [sic] as a co-author during the submission process, had their identity removed from the author by-line during the revision process by another co-author, without any documentation that this removed co-author had received notification and/or consented. This runs contrary to Committee on Publishing Ethics (COPE) position on good publishing practice in regard to changes in authorship (see http://publicationethics.org). This decision does not in any way reflect on the scientific findings presented in the article.

Why the long delay? The editor of the journal, John Quackenbush, referred us to Elsevier. A company spokesman told us that the person who lodged the complaint didn’t learn about the change in authorship until long after the paper was published.

Genomics decided it had to pull the article, with what strikes us as an odd caveat: If Camper could demonstrate, among other things, that she had properly notified Solomon and the editors of the authorship change, they would consider restoring the paper. The journal also wanted to wait until the investigation into Solomon’s conduct was complete — which again raises the question of why the retraction happened when it did. Was the de-retraction (is that better?) was meant to time with the release of the ORI report? Seems plausible, but then couldn’t the whole switcheroo been avoided entirely?

Camper describes the process with evident frustration, and we’re sympathetic. Camper told us that she explained to the editors that Solomon’s research had been expunged from the final version of the article. Therefore, authorship shouldn’t really be a question. Indeed, as readers of this blog are well aware, journals are quick to retract articles for including authors who didn’t play a role in a manuscript — they shouldn’t be in the business of doing the same to groups that try to weed out noncontributors.

To that point, Camper and her colleagues conducted their experiments again and are now confident in the science behind the paper as it now appears.

Others in the lab repeated all of the work before we resubmitted. I can stand behind all the work in the paper. Everything in that paper has been checked and double-checked.

The same cannot be said, however, for another paper of Solomon’s. In 2007, the Journal of Medical Genetics published a letter on which she was lead author that called do-over on a 2004 article she’d written as part of her doctoral work.

Here’s said letter, which in a bit of serendipity came out just as Camper was learning about the problems with Solomon’s work in her lab:

In our previous article,¹we used array‐comparative genomic hybridisation (CGH) analysis to identify Xq26–q27 duplications in families with X‐linked hypopituitarism (XH). The array‐CGH assay was validated using affected male genomic DNA from two previously characterised XH families that carry Xq26–q27 duplications^2,3 (fig 1B,C) and by interphase fluorescence in situ hybridisation (FISH) analysis (fig 2). Array‐CGH analysis of three novel XH families, A, B and C (fig 3), indicated that each of these contained a different Xq26–q27 duplication (fig 1D–G).

Recently, we repeated these array‐CGH experiments using a more extensive X chromosome array containing over 2000 bacterial artificial chromosome (BAC) clones.4 As expected, duplications were identified in males from the two previously characterised XH families.^2,3 However, we have been unable to detect Xq26–q27 duplications in affected males from families A, B and C. Furthermore, repeated quantitative real‐time PCR experiments performed by an independent collaborator in a blinded assay detected SOX3 duplication in affected males from control families,^2,3 but not in affected males from families A, B and C. On the basis of these new data, we now cannot replicate the observation of duplications in families A, B and C families that was published in our paper.

Meanwhile, Genomics and Elsevier will be posting the following notice next week, we’re told:

An allegation was made concerning a problem with co-authorship that upon receipt of further information appears to have been non-problematic, and therefore the retraction notice, which was in place for March 2011 through September 2011 has been lifted.

For Camper, the experiencing has been wrenching.

I have people depending on me for leading a research team. This is my first experience with this kind of thing and it’s been very difficult. I feel like the people who are getting hurt the most here are my lab and my colleagues.

As a footnote, Camper didn’t cut Solomon loose completely. The acknowledgments section of the newly restored article states:

We thank A.H. Mortensen, M.A. Potok, and N.M. Solomon for their contributions to this work.