The Journal of the American Chemical Society has issued an expression of concern over “the presentation of results” in a 2014 paper about a new probe for use in imaging.
We haven’t heard back from the journal nor the authors of the paper, so there’s not much we can officially say about what the journal is investigating.
Here’s the text from the expression of concern:
This publication is currently under editorial review. The journal has been alerted to concerns over the presentation of results described in this article. Its status will be updated upon completion of the editorial review.
The paper has been cited 12 times, according to Thomson Scientific’s Web of Knowledge.
One person who has raised concerns about the paper is Roger Wong, who used to work at Hong Kong University with last author Dan Yang. Wong told us his lab spent most of 2014 trying to replicate the paper:
We got almost all negative data. At a very late stage, after almost a year, we just wanted to take a look at the raw data to see if there’s something my entire team did wrong.
After Wong acquired the raw data for “Molecular Imaging of Peroxynitrite with HKGreen-4 in Live Cells and Tissues,” he began to suspect the authors were not being truthful about how they acquired the data, and that some of the images were composites from different experiments completed with different conditions.
Wong said he and his team reported their suspicions to the school. He also sent emails to JACS, on behalf of HKU’s Academic Staff Association.
Here are the concerns that Wong emailed to JACS:
- In Appendix 1, 4 examples showed that each published raw image are compiled by Maximum Intensity Projection from individual slice images at all and different z positions, and the complied image contains all characteristics of individual slice images. In other words, in order to have another perfectly identical compiled image (if the authors allege the existence of 26% images), the corresponding 5-7 individual z slice images would have also be perfectly identical. The odd is simply none.
- Like using different laser intensity, authors also used other different equipment parameters – different z slices (from 5-7 slices) that is corresponding to different z value (from 1.48um to 2.22um respectively) and compiled together for direct comparison in the composite figure panel. As shown in appendix 1, for instance the untreated negative control read 5 slices and 1.48 um z value and the E.coli positive control read 7 slices and 2.22 um.
- On page S25, authors stated that “Z-stack imaging was conducted with an oil lens by acquiring 5 consecutive photosections (0.37 um per section) … )” which was different from the 6 or 7 z slices actually used for compiling.
- In Appendix 2, raw images straight from raw data files with NO adjustment were compiled to make a side-by-side comparison with the published figure panel. First, images on both sides are found to be identical as no discernible difference was observed. Second, corresponding intensity levels on both sides are found to be in perfect match.
- If all images acquired at 26% as claimed by authors, E. coli and ABAH treatment in the published figure panel 4 would not be matching the intensity levels of the raw images when compared across the panel but at a much dimmer extent. However, since the intensity levels on both sides in Appendix2 match, the possibility of E.coli and ABAH were acquired at 26% laser intensity is NONE but rather acquired actually at 35%.
Wong forwarded us an email from Kenneth Leung, Associate Dean of Faculty of Sciences at HKU, which says that disciplinary committee (DC) investigated two students on the paper — second author Nai-Kei Wong and co-author Hu Jun — and cleared them of misconduct:
It was mainly because of lacking compelling evidences, and there were doubts on the charges (in particular cherry packing and laser intensity charges). For the charge on different dates, they believed that the students did the incorrect data presentation mainly due to negligence rather than deliberately making such a mistake; their admitted mistakes were not regarded as research misconduct by the DC.
However, the DC did recognize and identify there were bad research practices of the defendants/their laboratory, and will make recommendations to them, their supervisor, Department of Chemistry and Faculty of Science.
Wong, who says that Yang forced him to leave his lab at HKU, declined to share where he currently works.
We’ve reached out
to Yang, and to Leung. We’ll update this post with anything else we learn.
A spokesperson for HKU declined to comment:
In the light of the Personal Data (Privacy) Ordinance, it is not appropriate for the University to respond to or comment on the enquiries.
The editor in chief of JACS, Peter Stang, told us:
Concerns over the presentation of results described in this article were brought to the attention of the journal. The matter has been considered by the journal editors, and an Expression of Concern was posted alongside the article. The journal and the involved parties are working to resolve this matter. The status of the Expression of Concern associated with the article will be updated upon completion of the editorial review.
Update 3/9/16 4:31 p.m. eastern: When we were reporting this story, we tried to email Yang but just recently realized the email didn’t go through. We’ve now sent the email, and will update the story with any response. We apologize for this unfortunate oversight.
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy.