“I hold my hands up”: PubPeer criticisms spark a biochemist’s offer to retract

MPIn a back-and-forth on PubPeer, a researcher appears to have offered to retract a paper after commenters challenged the use of identical control bands on a gel.

A person writing as Virginia Commonwealth University biochemist Paul Dent admitted the control bands of his 2007 Molecular Pharmacology paper, “Human Chorionic Gonadotropin Modulates Prostate Cancer Cell Survival after Irradiation or HMG CoA Reductase Inhibitor Treatment,” were duplicated “for comparison purposes,” but stated he was “in no way attempting to ‘intentionally manipulate the data.’”

In response, Dent appeared to offer to pull the paper, about using a hormone to help kill prostate cancer cells:

If you wish me to write to The Journal and retract the entire manuscript, I will do so.

The paper has been cited seven times, according to Thomson Scientific’s Web of Knowledge. It is one of 16 of Dent’s papers that have sparked comment on PubPeer, some for seemingly similar bands in multiple gel lanes, others for “possible reuse” in different papers.

On May 17th, an anonymous PubPeer commenter pointed out distinctly straight lines on a gel in Figure 2D and seemingly identical bands in a gel in figure 2E of the Mol Pharm paper. Someone signing comments as Dent — and there is evidence, we should note, that this is in fact Dent — quickly responded to the criticism:

The data in Panel D was NOT cut and paste; that’s just the way the lanes ran. The vehicle values in Panel E are the same because the postdoc took a baseline time = 0 value. Which iwas used as the control value for both time points.

The anonymous poster responded:

30 minutes and 180 minutes after exposure means two samples. It is very unlikely that they would run so similarly.

If you ran your samples thousands of times it might happen, but this is in the next independent panel. This sort of thing happens quite often in your publications.

The person writing as Dent responded:

It is a mistake on my / our part to have labeled the figure in the way I did. There should be a bar between the vehicle value and the treatment values to indicate the vehicle control comparison is the same at both time points.

After another critic chimed in, the commenter writing as Dent offered to retract the paper:

As I recall the postdoc ran one vehicle control. That image was duplicated and the duplicate placed alongside the data from each time point for comparison purposes. I was the person who did that, I freely admit. However, I was in no way attempting to “intentionally manipulate the data”…. I simply duplicated the control value and placed it alongside the images of how the various treatments changed band intensity for the reader to more easily compare between control and treated. If that is classified truly as scientific misconduct then I hold my hands up. But that was in no way my intention. If you wish me to write to The Journal and retract the entire manuscript, I will do so.

The paper has yet to be retracted, and Dent has not responded to our efforts to reach him via email and phone.

In another PubPeer exchange, over a 2008 Molecular Pharmacology paper, someone commenting as Dent said he could not explain straight lines in a gel and offered to post his original data. He has yet to do so.

Hat tip: Leonid Schneider

Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.