The Japanese research institute RIKEN has come full circle in a way on the STAP cell scandal. Note that the STAP papers included not only authors from RIKEN, but also from Brigham and Women’s Hospital/Harvard Medical School.
With its final report released today (also a powerpoint of images were released including the one showing a figure posted here of reportedly made up data published in a STAP paper), RIKEN seems to now have handled this complicated mess in a relatively rigorous, scientific manner that paves the way for moving on from it.
What would you have done in RIKEN’s shoes? It’s not a pleasant thing to contemplate.
As a research institution, what do you do if a potential major scientific scandal is cropping up under your roof? It’s unpredictable and dangerous and couldn’t have come at a worse time for RIKEN. So how do you handle it? What would you have done?
RIKEN was faced with this kind of unpleasant reality early this year. A very high moment for RIKEN with the publication of two seemingly groundbreaking Nature papers including numerous RIKEN authors quickly headed the other direction and began unraveling.
By February, only weeks after the STAP cell papers were published supposedly reporting the creation of power stem cells called “STAP cells”, there were signs that the research was plagued with profound problems including signs of potential misconduct.
At that point in February and March, what should the leaders at RIKEN do/have done? I’m sure they were asking themselves this tough question.
There have been ups and downs as to how RIKEN has handled STAP, but in the new report there are indications of a sober, objective and more thorough approach to STAP. I think this is a positive, constructive step even if the conclusions are negative and sad.
Dennis Normile has a nice summary over at Science News of the RIKEN report:
“The committee determined that 3 supposed STAP stem cell lines were actually likely to be 3 previously existing embryonic stem (ES) cell lines. “It is unlikely that there was accidental contamination by three different ES cells, and it is suspected that the contamination may have occurred artificially…”
This new RIKEN report on STAP has concluded that what were claimed to be STAP cells were almost certainly embryonic stem cells (ESCs) instead. It cannot be sure if the ESCs were intentionally and fraudulently used, but it says that is probable. If it was done on purpose RIKEN also cannot be sure who did it.
The report concludes that STAP first author Haruko Obokata, who RIKEN earlier had determined committed certain STAP-related misconduct, committed additional misconduct over certain data, but the report also lays some of the blame for STAP to lack of proper supervision by senior STAP authors Yoshiki Sasai, Teruhiko Wakayama, and Hitoshi Niwa. One remaining thing we do not know is what this report means for Wakayama or Niwa, the latter still being at RIKEN.
So what does all of this mean and where do things go from here?
While there is still some more to resolve on the Japan side of STAP, my impression is that this report along with Obokata’s resignation allows for RIKEN to really begin to move on from STAP.
Normile’s article concludes pointing towards where the STAP-related focus may turn next:
One of the papers’ co-authors has been beyond the reach of RIKEN investigators: Charles Vacanti, a tissue engineer at Brigham and Women’s Hospital in Boston, an affiliate of Harvard Medical School. Obokata initiated her work on STAP cells while a post-doc in Vacanti’s lab. Mutsuhiro Arinobu, a RIKEN executive director, said that although they have been in contact with Harvard, input from Vacanti “is not included in this investigation.”
Harvard and Brigham and Women’s Hospital may or may not be conducting STAP investigations of their own. However, certainly at this point relatively speaking key unanswered questions remain on the Harvard side of STAP.
The most recent development was somewhat unexpected… At least to me.
* http://www.nikkei.com/article/DGXLASDG26HFT_W5A120C1CR8000/
* http://www.tokyo-sports.co.jp/nonsec/social/359470/
* many many Japanese sites spanning the range from Nikkei to Tokyo Sports 😉
(Above articles only in Japanese — so far, I haven’t seen this in any Western media)
So apparently a Riken retiree (Mr 石川智久, Tomohisa Ishikawa? Not sure of pronounciation…) has gone to the police with evidence that supposedly proves that H.O. stole the ES-cells from Wakayama lab that were later found to contaminate her supposedly STAP-containing samples.
Wow.
There’s going to be another full season of episodes of this soap opera on Japanese channels. Stay tuned.
There’s probably less to this than meets the eye.
The cops already know about the STAP misconduct case (as does everyone in Japan). If the cops had wanted to investigate they already would have done so on their own. The filing of the criminal complaint just forces the cops to go through the motions of investigating. They’ll send their investigation file to the prosecutors, who will presumably find there’s not enough evidence to indict. A citizen review panel may object, and the prosecutors will re-investigate and not indict again…… It’s highly likely that the whole thing will just fade out in the end.
Of course it goes without saying that if Riken’s finding of misconduct is correct (and since Obokata has not objected this finding is now finalized) there are reasonable grounds for suspecting misuse of public funds. (Guilt or innocence would in the end be adjudicated by the court in the event of prosecution.) IMO there ought to be a prosecution, but what ought to happen and what will happen often differ.
Robert, I agree with you on a few points. Not all. The main issue here is that a theft has been reported to the police. A theft is something the cops can’t investigate until someone tells them that something was stolen. Of course many other aspects the cops could have investigated already if they wanted… Moreover, the public nature of this mess makes it more complicated.
Under normal circumstances, the Japanese Police is reluctant to do anything about reported thefts, even if you bring them surveillance video that explicitly shows who stole what and when… But this is a high profile case, so the police may feel obliged to do something. For the same reason, the easiest way for the prosecution to avoid public bashing may be to take the case to court.
It may simply fade away, as you predict, or then proceed along an uglier path involving all possible courts etc. Either way, the press will love it.
——
Mr Ishikawa apparently filed the criminal complaint because he wants to clean Riken’s tainted reputation. Whatever course the police investigation (or lack of it) takes, the Japanese press is likely to follow this new branch, and give less and less attention to problems at Riken. Good move by Mr Ishikawa.
To HR:
Anytime!
Nowadays you need to put each author’s e-mail address during submission, don’t you? Then an automated message goes to each author, verifying authorship. I don’t know about Nature, but I believe most journals have implemented a system to make people stick to their policy. I bet these authors had given a chance to back down or do something right after submission.
can someone translate the japanese powerpoint slides?
I am trying to figure out what they are saying about the methylation analysis
Obokata submitted samples for the analysis to Riken’s sequencing core facility. They run the samples for her and reported back the results. She failed to provide any record of this experiment, but the facility has kept the original data. The committee tried to reconstruct her Nature figure from the original data.
She submitted 96 clones (shown on the right). Among those 96, only 3 clones had less than 3 metylated sites out of 11 CG pairs, yet her Nature figure reported 18 such low metylated clones (shown in red). Likewise, the original data contained only 4 clones of more than 9 metylated sites, yet her Nature figure contained 17 such highly metylated clones (shown in green). The comittie concluded that she fabricated the figure because it was impossible to reconstruct her figures from raw data.
Some of the middle authors are staff of this sequencing facility. I hope this partially answers Sachi’s question.
Ken, I’ve been staff at several research facilities, but that doesn’t put me as a middle author of papers submitted to Nature (or anywhere else).
http://www.nature.com/authors/policies/authorship.html tells us that:
… Submission to a Nature journal is taken by the journal to mean that all the listed authors have agreed all of the contents, including the author list and author contributions statements. … The author list should include all appropriate researchers and no others. Authorship provides credit for a researcher’s contributions to a study and carries accountability. …
It’s quite clear that every author is accountable for what is written in a manuscript submitted for publication.
That’s fascinating, thanks for the translation.
But this raises the question of why the sequencing facility staff were willing to put their name on a paper that contained inaccurate representations of results from their own facility.
Either they never saw the true raw sequencing results (in which case, why were they on the paper?); or they did see them but didn’t notice the discrepancy with the paper; or they did notice and said nothing…
I’m just fed up with a huge discussion on authorship with my collaborator and don’t feel like going into it again any time soon, but it’s fair to say that everyone seems to have his/her own sense and definition for authorship. I myself have included a tech from a virus core in my manuscript because her contribution was critical for the study.
A couple of clarification:
The said facility helped Obokata for ChIP-seq and RNA-seq as well.
The two were not among the original authors and were added at some point later during resubmission/revision.
These authors have publication as middle author with other people in Riken, indicating the Nature paper was not an anomaly for Riken.
I don’t think these people’s contribution was just loading Obokata’s samples into the machine. Rather, they most likely helped her to design the experiments and analyze the data because genome analysis was not her expertise. To what extent? I don’t know, but apparently, they were not informed which sample corresponds to which condition and thus were not in a position to assess the accuracy of the Nature figure in question. (She submitted three sets of samples for DNA methylation. At least they could tell which set was for the Oct4 promoter and which set was for the Nanog promoter.) Again, one may argue that these circumstances disqualifies these two from the author list and even that they should do so voluntarily, but I am yet to meet a person who is NOT willing to be a Nature paper author.
Hi Ken:
I’m one of those people who was NOT willing to be a Nature paper author.
In this case, the senior author would not let me see the paper or analyze the data to be published- hence, no authorship for me.
My reputation is more important to me than being on a Nature paper that I can’t vouch for.
Jeanne
Jeanne, I’m glad to hear that.
To those who are intersted, Riken’s report dated June 10, 2014 detailed each author’s contribution except the Harvard group. It says all the non-senior authors and the director of the EM core were not given a chance to review the manuscripts before submission or revision. I believe the report is not translated, but the page 19 may be useful to some:
http://www3.riken.jp/stap/j/c13document14.pdf
小保方: Obokata; 若山: Wakayama; 笹井: Sasai; 丹羽: Niwa; バカンティ: Vacanti; 小島: Kojima; 米村: Yonemura; 大和: Yamato; 共著者: coauthor
Thanks Ken!
I had a look at the Riken report, and it looks like I need to read it with care 😉
I once again quote Nature’s authorship policy: “Submission to a Nature journal is taken by the journal to mean that all the listed authors have agreed all of the contents”… I wonder how this agreement can be achieved if all authors are not given a chance to review the manuscript?!?
I would like to get a comment from Nature regarding such negligence of their authorship policies. After all, their policies are useless if nobody adheres to them…
As we all know, Nature requires authors to state who did and what.
* The article lists author contributions as follows: H.O. and Y.S. wrote the manuscript. H.O., T.W. and Y.S. performed experiments, and K.K. assisted with H.O.’s transplantation experiments. H.O., T.W., Y.S., H.N. and C.A.V. designed the project. M.P.V. and M.Y. helped with the design and evaluation of the project.
* The letter lists author contributions as follows: H.O. and Y.S. wrote the manuscript. H.O., Y.S., M.K., M.A., N.T., S.Y. and T.W. performed experiments, and M.T. and Y.T. assisted with H.O.’s experiments. H.O., Y.S., H.N., C.A.V. and T.W. designed the project.
The biggest responsibility clearly lies in the hands of whoever wrote the paper, and Y.S.’s role will never see the light of day (I’m not suggesting Y.S. had any role in the scam, but if he did, it would never come out. He was not investigated by Katsura’s panel, and won’t be investigated by others, either.)
Some authors are given the special status to have ‘designed the project’ or even ‘evaluated’ it. RIKEN investigations seem to clear T.W. and H.N., but I’d say also C.A.V., M.P.V. and M.Y. deserve some attention.
In papers with this many authors, it’s fair to say that whoever ‘performed experiments’ using samples provided by the project leader probably was doing so ‘in good faith’, and probably is as upset with the fuss as the rest of the scientific community. However, it should be clarified what was the role of those (K.K., M.T., and Y.T.) who ‘assisted with H.O.’s experiments’. How closely did they work with H.O., and why didn’t they notice anything fishy?
And then of course we have H.O.’s mentors and supervisors from Waseda and TWMU who should have quite good an insight of what really happened. At least they should know whether or not H.O. ever knew something about ‘scientific integrity’ — if not, they shouldn’t supervise or mentor any student ever again… These are the people who (most likely) helped her to get a position at RIKEN (and Harvard) to start with…
Paul, I think that you have given a free hand to quite a number of co-authors who were credited in the two retracted papers in Nature (July 2, 2014). The first paper (article) had 8 authors (Obokata, Wakayama, Sasai, Kojima, Martin Vacanti, Niwa, Yamato and Charles Vacanti). The second paper (letter) had 11 authors (Obokata, Sasai, Niwa, Kadota, Andrabi, Takata, Tokoro, Terashita, Yonemura, Charles Vacanti and Wakayama).
The attention for the STAP fiasco have been focused only on Obokata, Wakayama, Sasai, Niwa and Charles Vacanti. Why other authors have been silent so far? You had mentioned that you attempted to interview Obokata, but failed. How about approaching many other co-authors in the second paper (letter), whose corresponding address is RIKEN (Kobe). They should be knowing more about this sordid episode. Have their mouths been silenced?
“why other authors have been silent so far?”
I don’t know the specifics of the case, but I can imagine a lot of scenarios where the first author gives co-authors reagents to do specific experiments on– for example, the first author gives a cell pellet to a middle author to do western blots/RT-PCR/microscopy/etc…
The middle-author then goes about the characterization ignorant of what cells he or she was actually given. In this case, there is not much for the middle author to say…
We can analyze and debate what happened at RIKEN for as long as want. But we have no way of knowing what happened at Harvard, absolutely nothing. Vacanti was senior author. By rights he should have been the one fielding the questions! Especially since he seems to still believe in STAP. Why isn’t he defending Obokata?
Vacanti was, at least nominally, on the committee that accepted her doctoral dissertation, which was later found to be riddled with problems such as outright plagiarism and fabrication. There were reports that he had never read the paper in the first place. I just wonder if this (i.e agreeing to grant a doctorate on the basis of a paper one hasn’t read) is an accepted practice in biology or whatever field they are in.
Thank you for your in-depth coverage. Yes, it’s hard to contemplate to be in RIKEN’s shoes. I’m sure I will never be in Obokata’s shoes, but what about Wakayama, Tanba, or Sasai? Aren’t we, modern biomedical scientists, wearing their shoes every day?
Some people blame these senior authors for not asking for Obokata’s lab notebook or checking her raw data. I found this suggestion a bit too unrealistic. I routinely collaborate with other labs, but never asked for a lab notebook or raw data unless I needed to analyze it myself for a different purpose.
I found the RIKEN’s final report that you linked very insightful. There were several precious moments when these senior authors could have taken a different step. A Nature Protocol Exchange that followed the original papers included a sentence that TCR rearrangements were not observed in so-called STAP stem cells, which contradicted their data with so-called STAP cells in their initial report. This inconsistency was shared by the authors and bothered Tanba enough to suggest Sasai not to include this data in the manuscript. Sasai countered that STAP cells were heterogeneous, thus lost this trait through numerous passages. STAP stem cells were supposedly derived from mature T cells and it was a moment for Tanba to question its origin more rigorously, which he failed.
Another instance was with Wakayama. He back-crossed a 4N chimera with a wildtype and obtained GFP-negative pups, which was supposedly impossible because the STAP stem cells used to make the chimera was GFP(+/+). He and Obokata didn’t pursue this apparent discrepancy any further. The point is that Obokata didn’t manage mouse colony at all and it was Wakayama who did this mating alone by himself. He testified that he thought he was the one who made a mistake and that when he helps somebody else’s project and observes some discrepancy, he doubts himself first, but not the other person.
Really? An established researcher such as Wakayama questions himself before questioning a novice research such as Obokata? I may be a bit too egoistic, but I always assume that I am the one who is doing right. I always assume that the reason I don’t get data consistent with somebody else is because the person was not doing it right, even if the data was from a prestigious lab. Or at least I try to understand how and why the discrepancy happened. We, scientist, are trained to look at our own and somebody’s data critically and it seems that was all it needed to avoid this problem.