A short seller who is betting against the stock has raised image-duplication claims about a Nobel laureate's papers

What happened

On 7 August, Gotham City Research LLC, a short-selling firm based in the United States, published a report alleging image manipulation across papers published between 2000 and 2020 on which Shinya Yamanaka was an author. Yamanaka shared the 2012 Nobel Prize in Physiology or Medicine for the discovery of induced pluripotent stem cells, and directed Kyoto University's Center for iPS Cell Research and Application (CiRA) from 2008 to 2022.

CiRA issued a statement rejecting the report. As Retraction Watch reported on 27 August, the institute's position is that the report treats cases of very different standing as though they were one thing:

"Treating these different cases collectively as though they demonstrate an ongoing pattern of research misconduct by Professor Yamanaka or by CiRA disregards the facts specific to each case and is completely unacceptable."

Hiroyuki Wadahama of CiRA's international public communications office told Retraction Watch the institute is not aware of any ongoing institutional investigation into the papers, and that CiRA had no contact from Gotham before publication: "We have not been in touch with Gotham City Research regarding their report, nor did we receive inquiries from them prior to its publication."

The financial position, stated plainly

The report's nominal targets are two Japanese companies, Sumitomo Chemical and Sumitomo Pharmaceutical, both of which have funded CiRA. Gotham's founder Daniel Yu told Retraction Watch the firm has shorted the stock of both, meaning it profits if the share prices fall.

This is disclosed rather than hidden, and short sellers have raised research-integrity questions before, sometimes with substance behind them. In 2021 a group publicly doubted Cassava Sciences' experimental Alzheimer's drug, which later failed in clinical trials. Cassava sued the short sellers and dropped the defamation suit in 2024. Gotham's own 2024 campaign against Grifols SA removed roughly a third of that company's value and was followed by a governance restructuring.

What the ten papers actually consist of

This is the part worth reading slowly, because it is where the dispute is really located. According to CiRA's account as reported by Retraction Watch, the ten papers break down as follows.

• Yamanaka was first or corresponding author on three.
• On four, he is a coauthor because he supplied research materials such as cells or mice.
• On two, he is not an author at all. One was retracted by Stem Cell Reports in 2018 after Kyoto University found a colleague, Kohei Yamamizu, had fabricated and falsified data, and dismissed him. A second paper listing Yamamizu was retracted in 2019. The university's investigation found Yamamizu solely responsible; Yamanaka said at the time that he felt "a strong responsibility for not having prevented research misconduct at our institute" and was himself penalised by CiRA.
• Of the three where he is first or corresponding author, two have already been through a process. A 2008 Science paper was corrected. For a 2000 paper in The EMBO Journal, the editors said in 2014 they saw "no concrete evidence" of duplication. For the third, a 2003 paper in Cell, CiRA says Yamanaka reviewed the images in response to this report and concluded the two alleged to be similar "are not identical".
• A further paper, published in Human Molecular Genetics in 2011, was corrected last year. The correction notice records that the raw data were not fully available but that the authors determined inadvertent errors occurred during final figure assembly.
• On a 2020 paper in the Journal of Neuroscience Research, corresponding author Jun Takahashi, who directed CiRA from April 2022 to March 2026, is reviewing the concerns and will take "appropriate action".

Most of the flagged images were already posted on PubPeer, the post-publication review site, before the report appeared.

Analysis

The following is Bizarus analysis, not reporting. Nothing here should be read as a finding about any paper or any researcher.

Two things are true at once and they are easy to collapse into each other.

The first is that a financial interest does not make evidence false. Whether two panels in a figure are the same image is a question about the figure. It is answerable by anyone with the file, it does not become a different question depending on who asks it, and dismissing a duplication claim because of who raised it would be exactly the reasoning that lets misconduct survive scrutiny from inconvenient sources. Sleuths working outside institutions have been right often enough that motive is a poor filter.

The second is that a financial interest changes what gets selected and how it gets assembled, and that is where CiRA's objection actually bites. Nobody chooses ten papers at random. A claim that ten items constitute a pattern is a claim over and above any claim about an individual image, and it needs its own support: that the items are comparable, that they are not already resolved, and that the named person bears a comparable relationship to each. On CiRA's account, this set contains two papers Yamanaka did not write, four where he supplied mice or cells, one already corrected, one where journal editors found no evidence a decade ago, and one corrected last year with an explanation on the record. A reader who sees only the number ten sees something quite different from a reader who sees the composition.

That move, aggregating items of unlike status into a single number and letting the number carry an implication none of the items carries alone, is not specific to short sellers. It is one of the ordinary ways a defensible set of observations becomes an indefensible conclusion, and it is worth recognising whoever is doing it, including when the aggregator is someone whose conclusion you find congenial.

What is genuinely new here is the configuration. Post-publication scrutiny has been drifting outside journals and institutions for a decade, toward PubPeer commenters and individual sleuths, mostly unpaid. A firm with a disclosed position in the share price is a different kind of outside party: better resourced, more visible, and with a reason to prefer one answer. That does not disqualify it. It does mean the incentive belongs in the reader's field of view alongside the images, which is presumably why Gotham disclosed it.

What to watch

Whether Kyoto University or any journal opens a formal inquiry, since CiRA says none is currently under way, and what Jun Takahashi concludes on the 2020 paper. Separately, whether the papers behind Amchepry, the reprogrammed-stem-cell treatment for Parkinson's disease that Japan's health ministry conditionally approved this year and which has so far been tested in seven patients, are drawn into the dispute. They have not been so far.

Sources

• Dalmeet Singh Chawla, Japan institute defends Nobel Prize winner amid allegations of image duplication, Retraction Watch, 27 August 2026. CiRA's statement and Daniel Yu's confirmation of the short position are quoted as reported there; the CiRA press-release page could not be opened directly from this environment for independent verification.
• Gotham City Research, Sumitomo Chemical and Sumitomo Pharmaceutical, Part II, 7 August 2026. The report is an advocacy document published by a party holding a short position in the companies it discusses.
• Retraction Watch, Stem cell paper falsification leads to firing; Nobelist also penalized, 29 March 2018.