UCSF-VA investigation finds misconduct in highly cited PNAS paper

PNAS has corrected a highly cited paper after an investigation found evidence of misconduct.

The investigationconducted jointly by the University of California, San Francisco, and the San Francisco Veterans Administration Medical Centeruncovered image manipulation in Figure 2D, which “could only have occurred intentionally.” The institutions, however, could not definitively attribute the research misconduct to any individual.

According to the notice, the UCSF-VA committee determined that a correction to the 2008 PNAS paperwhich explores the genetic underpinnings of prostate cancerwas “appropriate,” and the authors have now replaced the problematic figure with a corrected version. The 2008 paper has been cited 630 times, according to Clarivate Analytics’ Web of Science.

A spokesperson for PNAS told Retraction Watch:

The Committee determined that a correction including the original data was appropriate in this case. … We have nothing further to add to the Correction notice.

But is a correction really appropriate when an investigation finds evidence of misconduct?

Recently, we’ve noted a few cases where journals have opted to retract papers because of a UCSF-VA investigation, which we wrote about last year.

Three papers were retracted last year after a UCSF-VA committee sent letters to the American Association of Cancer Research (AACR) — which publishes Cancer Research and Clinical Cancer Research — and to the International Journal of Cancer. According to the letters, the institutions found “clear evidence” of research misconduct, but left it up to the journals to decide whether to correct or retract the papers. All three journals chose retraction.

We suspect the UCSF-VA investigation mentioned by the previous three retractions is the same one referenced in the latest correction, as all papers share the same last author: Rajvir Dahiya, director of the Urology Research Center at UCSF. We have contacted the institutions to verify whether the PNAS paper was part of the joint investigation cited by the previous three retractions, and whether we can expect more retractions and corrections. A representative of the VA referred us to UCSF and the journal.

The institutions could not determine who was responsible for the misconduct; the authors said the figures had been created by two researchers who had left the country and couldn’t be contacted. Neither of the two researchers are co-authors of the PNAS paper.  

Here’s the correction for the 2008 paper, “MicroRNA-373 induces expression of genes with complementary promoter sequences:”

The authors wish to note the following: “The corresponding authors were made aware of errors in Fig. 2D that required further investigation. The University of California, San Francisco (UCSF) and VA Medical Center, San Francisco, conducted a joint investigation into the cause of the errors. The investigation was overseen by the VA Office of Research Oversight (ORO) and, in part, by the UCSF Office of Research Integrity (ORI) independent of the authors.

“The Investigation Committee reviewed Fig. 2D including the E-cadherin panel (lanes 5 and 6) and GAPDH panel (lanes 2 and 5) and concluded that these images were derived from the same source of data despite representing different experimental conditions. Further analysis also led to the finding that a portion of the GAPDH panel, encompassing approximately 1.5 lanes, was mirrored and added to the panel. The Committee concluded that the manipulation of the images in Fig. 2D could only have occurred intentionally, representing instances of scientific misconduct. The Committee could not definitively attribute the research misconduct to any individual.

“One author then was able to identify an image copy of the original GAPDH immunoblot and two exposures of a corresponding E-cadherin immunoblot in their digital archives. The data were then submitted to the Committee and PNAS for evaluation. After review, it was determined by the Committee that a correction to Fig. 2D would be appropriate. The correction does not impact interpretation of the data.”

The 2008 paper was funded in part by $3.6 million in National Institutes of Health grants awarded to Dahiya.

We contacted Dahiya and the other corresponding author, Robert F. Place, who was based at the VA and UCSF but has a more recent affiliation at the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Maryland.

Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.

9 thoughts on “UCSF-VA investigation finds misconduct in highly cited PNAS paper”

  1. Oncogene. 2007 Dec 6;26(55):7647-55. Epub 2007 Jun 11.
    Knockdown of astrocyte-elevated gene-1 inhibits prostate cancer progression through upregulation of FOXO3a activity.
    Kikuno N1, Shiina H, Urakami S, Kawamoto K, Hirata H, Tanaka Y, Place RF, Pookot D, Majid S, Igawa M, Dahiya R.
    Author information

    Department of Urology, Veterans Affairs Medical Center and University of California, San Francisco, CA 94121, USA.


  2. If we were investigating people whose tax returns were incorrect, and every time the outcome was “The return was incorrect, but correcting it does not increase the amount of tax owed”–would that be plausible? Presumably people falsify their tax returns in the hope of paying less taxes.

    It seems likely to me that, while sometimes people will falsify their blots just to save time and effort, it generally is done to make the paper more attractive and publishable. That being the case, I’m deeply suspicious of the parade of “This does not affect the conclusions” that we get along with corrections. In some cases it may be true, but across the board??

    1. I agree with Dr. Kuhner. Further, when you fraudulently prepare and submit your tax forms, is the standard IRS response limited to having you agree to a three-year period of increased oversight?

  3. So what should I tell our students? That research misconduct is okay?!
    This is really devastating news.

    Research institutions are blinded by the overhead from research grants and apparently are willing to accept everything. Some journals have understood this problematic conflict of interest., but not PNAS.

  4. > the authors said the figures had been created by two researchers who had left the country and couldn’t be contacted. Neither of the two researchers are co-authors of the PNAS paper.

    Alright. I guess their names were John Doe and Richard Roe. It is amazing how people can get away with misconduct these days.

  5. Oncotarget. 2017 Jun 13;8(24):39087-39100. doi: 10.18632/oncotarget.16598.
    Cytochrome P450 1B1 inhibition suppresses tumorigenicity of prostate cancer via caspase-1 activation.
    Chang I1, Mitsui Y2,3, Kim SK1,4, Sun JS1,4, Jeon HS1, Kang JY1,4, Kang NJ1,4, Fukuhara S2,3, Gill A2, Shahryari V2, Tabatabai ZL5, Greene KL5, Dahiya R2,3, Shin DM1,4, Tanaka Y2,3.
    Author information
    Department of Oral Biology, Yonsei University College of Dentistry, Seoul, South Korea.
    Department of Surgery and Division of Urology, Veterans Affairs Medical Center, San Francisco, California, United States of America.
    Department of Urology, University of California, San Francisco, California, United States of America.
    BK21 PLUS Project, Yonsei University College of Dentistry, Seoul, South Korea.
    Department of Pathology, Veterans Affairs Medical Center and University of California, San Francisco, California, United States of America.

    See: https://pubpeer.com/publications/2AC631A3905D2F944FF0D403AAB0C9

    2018 correction. http://www.oncotarget.com/index.php?journal=oncotarget&page=article&op=view&path%5B%5D=26197&pubmed-linkout=1

    This article has been corrected: The image for #4-3 in Figure 2C is incorrect. It is a duplicate image of panel #4-2. The corrected Figure 2C is shown below. The authors declare that these corrections do not change the results or conclusions of this paper.

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.