Researchers Plan To Retract Landmark Alzheimer's Paper Containing Doctored Images (science.org) 69
An anonymous reader quotes a report from Science Magazine: Authors of a landmark Alzheimer's disease research paper published in Nature in 2006 have agreed to retract the study in response to allegations of image manipulation. University of Minnesota (UMN) Twin Cities neuroscientist Karen Ashe, the paper's senior author, acknowledged in a post on the journal discussion site PubPeer that the paper contains doctored images. The study has been cited nearly 2500 times, and would be the most cited paper ever to be retracted, according to Retraction Watch data. "Although I had no knowledge of any image manipulations in the published paper until it was brought to my attention two years ago," Ashe wrote on PubPeer, "it is clear that several of the figures in Lesne et al. (2006) have been manipulated ... for which I as the senior and corresponding author take ultimate responsibility." After initially arguing the paper's problems could be addressed with a correction, Ashe said in another post last week that all of the authors had agreed to a retraction -- with the exception of its first author, UMN neuro-scientist Sylvain Lesne, a protege of Ashe's who was the focus of a 2022 investigation by Science. "It's unfortunate that it has taken 2 years to make the decision to retract," says Donna Wilcock, an Indiana University neuroscientist and editor of the journal Alzheimer's & Dementia. "The evidence of manipulation was overwhelming."
The 2006 paper suggested an amyloid beta (AB) protein called AB*56 could cause Alzheimer's. AB proteins have long been linked to the disease. The authors reported that AB*56 was present in mice genetically engineered to develop an Alzheimer's-like condition, and that it built up in step with their cognitive decline. The team also reported memory deficits in rats injected with AB*56. For years researchers had tried to improve Alzheimer's outcomes by stripping amyloid proteins from the brain, but the experimental drugs all failed. AB*56 seemed to offer a more specific and promising therapeutic target, and many embraced the finding. Funding for related work rose sharply. But the Science investigation revealed evidence that the Nature paper and numerous others co-authored by Lesne, some listing Ashe as senior author, appeared to use manipulated data. After the story was published, leading scientists who had cited the paper to support their own experiments questioned whether AB*56 could be reliably detected and purified as described by Lesne and Ashe -- or even existed. Some said the problems in that paper and others supported fresh doubts about the dominant hypothesis that amyloid drives Alzheimer's. Others maintained that the hypothesis remains viable. That debate has continued amid the approval of the antiamyloid drug Leqembi, which modestly slows cognitive decline but carries risks of serious or even fatal brain swelling or bleeding.
The 2006 paper suggested an amyloid beta (AB) protein called AB*56 could cause Alzheimer's. AB proteins have long been linked to the disease. The authors reported that AB*56 was present in mice genetically engineered to develop an Alzheimer's-like condition, and that it built up in step with their cognitive decline. The team also reported memory deficits in rats injected with AB*56. For years researchers had tried to improve Alzheimer's outcomes by stripping amyloid proteins from the brain, but the experimental drugs all failed. AB*56 seemed to offer a more specific and promising therapeutic target, and many embraced the finding. Funding for related work rose sharply. But the Science investigation revealed evidence that the Nature paper and numerous others co-authored by Lesne, some listing Ashe as senior author, appeared to use manipulated data. After the story was published, leading scientists who had cited the paper to support their own experiments questioned whether AB*56 could be reliably detected and purified as described by Lesne and Ashe -- or even existed. Some said the problems in that paper and others supported fresh doubts about the dominant hypothesis that amyloid drives Alzheimer's. Others maintained that the hypothesis remains viable. That debate has continued amid the approval of the antiamyloid drug Leqembi, which modestly slows cognitive decline but carries risks of serious or even fatal brain swelling or bleeding.
Science working as it should (Score:4, Interesting)
Look, I won't sugar coat this message. A mistake of this magnitude is unforgivable to the scientists who had knowledge of it. The senior author is acting appropriately. Science is working as intended, recognizing mistakes and correcting them.
The alternative world is where you believe everything that is put forth. That path leads to far worse outcomes.
Strongly disagree (Score:3, Interesting)
Strongly disagree.
Publishing is not the same as science. going on 20 years - that is a whole generation of research largely going in the worng direction. Billiions of dollars. 100s of millions dying bereft of dignity.
If this were science as it should be then replication would be as important as the original research. But that is not the way the game is played.
F*** these guys - hanging is not enough for what they did. Instead they got fame and riches.
Re: (Score:3, Insightful)
Strongly disagree.
Publishing is not the same as science. going on 20 years - that is a whole generation of research largely going in the worng direction. Billiions of dollars. 100s of millions dying bereft of dignity.
If this were science as it should be then replication would be as important as the original research. But that is not the way the game is played.
F*** these guys - hanging is not enough for what they did. Instead they got fame and riches.
Clearly written by a non-scientist. If you do research and do not disseminate it, you have done nothing to advance humanity. Publishing is a critical part of science. Sharing your results with others, teaching, is baked into what we, as scientists, do.
The false dichotomy that without this paper hundreds of millions of people would have died with dignity is laughable in its exaggeration.
Hanging not good enough? You want to torture and kill a talented team because one of them was unscrupulous? Please.
Thi
Re:Strongly disagree (Score:5, Insightful)
Publishing is a critical part of science.
It is, but science publishing has a strong bias towards new positive results.
It is difficult for scientists to publish negative results and even harder to publish replication research.
The result is that researchers avoid risky research that is likely to fail but could lead to a big breakthrough if successful.
They have even less incentive to do replication.
Since there is pressure to get a positive result, a negative result is faked or p-hacked into a positive result, which is often successful since no one is doing replication.
This is science working as it should: discovering the mistakes and correcting them.
It took years for the fudgery to be exposed and more years for it to be retracted. This was an extremely high-profile paper. Similar fakery in an obscure paper would likely never be noticed.
Re: (Score:2, Flamebait)
Don't blame scientists for lack of negative results papers. Blame yourself for not funding scientists and labs to do replication studies and negative result publishing. How would a negative results journal even work (they already exist btw, or were tried in the past and failed): https://jnrbm.biomedcentral.co... [biomedcentral.com] ) "Checked under the planet and didn't find turtles holding it up." Yeah there's fame and fortune for that.
Re:Strongly disagree (Score:5, Interesting)
Don't blame scientists for lack of negative results papers.
Scientists decide which journals to read and cite.
Scientists do the peer reviews that determine which papers are published in those journals.
Scientists make the hiring and tenure decisions for other scientists based on the publication records in those journals.
It's scientists all the way down.
Blame yourself for not funding scientists
So scientists are mismanaging the money they have, thus the solution is to give them more, with no change in the process or accountability? Why would anyone believe that won't lead to more of the same?
Re: (Score:2)
So scientists are mismanaging the money they have, thus the solution is to give them more, with no change in the process or accountability?
No. You clearly don't know anything about scientists or science funding.
Scientists don't stump up the money, and scientists don't award people with jobs. Those two groups want to see "results", which means new, interesting things which people care about, and ideally get public buzz. No one remembers the names of the second guys to land on the moon.
If scientists some how
Re: (Score:2, Troll)
Uh, you said he's wrong then wrote a detailed explanation of why he's right....
Re: (Score:2, Flamebait)
Uh, no I didn't.
The problem is structural, and relentlessly ejects scientists from positions of influence that do what you want. And yet somehow you've figured it's the scientists fault collectively.
Then again simplifying a massively complex, nuanced and dynamic issue to being entirely the collective fault of an "other" group is about the level of discourse I'd expect from either of you.
Re: (Score:2)
I'm only noting your logical and rhetorical error. I don't have a take on the topic.
I read and edited my friend's college papers for free because I enjoyed it.
Re: (Score:1)
I'm only noting your logical and rhetorical error
lol nope.
I read and edited my friend's college papers for free because I enjoyed it.
Please convey my sympathies to your friend.
Re: (Score:2)
My friends bought me beer for improving their papers.
You get my service entirely for free. Although not bright enough to recognize your poor logic and rhetorical failures so it is a case of pearls before swine, but that's not uncommon.
Re: (Score:2)
My friends bought me beer for improving their papers.
Sure they did. Them everybody clapped.
You get my service entirely for free.
Well I'd say your services are seriously overpriced.
Especially as you can't come up with the goods, only assure me that they exist. ;)
Re: (Score:2)
So scientists are mismanaging the money they have,
Scientists have to commit to mismanaging the money before they are given the money. Science is one of the most profitable things there is, but the profits don't go to the scientists, so science is funded by external sources. And those external sources insist on atrocious measures when deciding who to fund (publish or perish type stuff, never do replication studies, etc -- beg us for grant money).
The world would be a vastly different place if the profits generated by the use of Newtonian mechanics, quantum m
Re: (Score:2)
"It is, but science publishing has a strong bias towards new positive results."
It doesn't. This is a misunderstanding of statistics. Publishing has a strong bias towards results, as it should.
*Scientists* (in some fields) have a strong bias towards positive results because the little bit of statistics they're taught doesn't tell them how to produce negative results. p > 0.05 is not a negative result, it's an inconclusive one that shouldn't be published. In order to produce an actual negative result you h
Re: (Score:3)
Actually, it does have that kind of bias. There are LOTS of papers that never get published, or get published in journals that nobody reads.
Scientists, in every field that I am aware of, have a bias in favor of interesting results. But their bias is minor compared to that of the journal publishers. (And the publishers have a different idea of what "interesting" means.) So if you do a paper that produces expected results, you can't expect to get published in a prestige journal. And perhaps not at all.
Re: (Score:2)
Negative results are interesting. You'd said nothing to indicate that you're not just talking about inconclusive results, like everyone else, which are not interesting.
It's very easy for people to claim there's a positive publication bias. It puts the blame on publishers, never mind that the ones rejecting the papers are the reviewers, and the scientists don't need to learn anything new, do any more work, or actually power their experiments appropriately. Funding agencies don't need to pay for any of that e
Re: (Score:3, Informative)
This was fraud. Fraud to get additional funding, fraud to further their careers, fraud. Plain and simple.
We're not talking about an honest mistake, we're talking about at least one of the authors purposefully, and willingly manipulating images with an agenda. We're talking about others potentially knowing about it, or not policing the actions of their peer group.
When fraud, when purposeful dishonesty is involved, at the very least this should enter a court, similar to criminal court, naturally with a jud
Re: (Score:2)
What's happening here is that there has been a MASSIVE uptick in this sort of behaviour.
And what drives the behaviour, eh?
The system is set up to prey on the desperate. Academia is absolutely brutal, for up and coming academics. Gaol is not going to be a disincentive here because no one thinks they will get caught. People are already wrapped up completely in a system they cannot see a way out of.
All you will do with your proposal is increase the prison population slightly at great expense. It won't reduce
Re: (Score:2)
"And what drives the behaviour, eh?"
Irrelevant and meaningless.
You certainly do your best to reduce the number of crimes, by doing what can be done to reduce conditions to create that situation. That has nothing to do with using criminal law to curb sociopathic behaviour.
This isn't a case of someone stealing bread to feed a starving family. This is a case of fraud for gain, at a cost to others, a morally reprehensible act. In fact, under current existing law, if this was done to obtain a grant, or to fur
Re: (Score:1)
Irrelevant and meaningless.
I mean yes if you are the "law and order" type who prefers retribution over reducing crime, then yes it's irrelevant and meaningless. If you prefer to have things be better then no.
Don't pretend that external circumstances have zero difference on behaviour.
This is a case of fraud for gain, at a cost to others, a morally reprehensible act./em.
Yes, and? If you set the system up to systematically eliminate all the careful scientists, and keep only the most relentless career builders,
Re: (Score:2)
*Don't pretend that external circumstances have zero difference on behaviour.*
You missed the entire next sentence, which states that migratory methods may be employed as well. Your response, where you purposefully ignored that fact, shows you are debating without good faith.
And yes, you still need to have criminal acts dealt with. Are you next going to suggest that insider trading shouldn't be illegal, or punishable in criminal court, because "Those poor traders were raised wrong" or "The system encourage
Re: (Score:2)
Well - I have a PhD in MIS, another ABD in stat, an MBA, a MS in EE, and a BS in Chem E. I would put my "science" credentials against yours any day of the week. Again, imo, our current "science" publishing schema is a perversion, with an emphasis on wiriting papers with statistical significance lacking verification and in most cases pracitical significance.
And don't even get me started on how bad most faculty are at teaching -- which is why most parents even care about academics in the first place. I was
Re: (Score:3)
Most of your points are valid, but this is *NOT* science working as it should. Were science working as it should, replication would have immediately be a high priority on several people's agendas. As it is, if they had done a replication, they'd have gotten a pat on the head from their peers, and a kick in the ass from their departments.
This is merely "the best we can do in the current environment". It's a lot better than hiding and obfuscating, but it's not "things working as they should".
Refunds for journal subscriptions (Score:2)
Nature should provide refunds for the journal subscription for the issue containing the falsified data.
A $240 subscription would need to refund 1/12 or $20 to each subscriber for that year.
Re: (Score:2)
It would cost way more than $20 per subscriber to verify any non-theoretical the papers that are selected for publication. If someone fakes a chart, assuming it doesn't present a logical fallacy, it is nearly impossible to verify it without actually carrying out the experiment which would be very costly and time-consuming.
That's the idea (Score:2)
If there are minor consequences for academic journals to publish and retract falsified research articles, there is no incentive to correct the larger process of research - submit - peer review - journal publish process.
Larger issue, can government grants be withheld, student loan money withheld from entire universities which routinely publish falsified research?
The threat of withholding federal funds has been made by the federal government before.
Re: (Score:2)
So you want to incentivise Nature to not request article retractions?
I think that's called a malign incentive.
Re: (Score:3)
Huh? I don't think anyone died from this .. did any AB*56 targeting drugs reach market or even clinical testing phase? As for research money .. I doubt billions or even 100 millions were spent on AB*56 research. It seems highly improbable that so much got spent on something nobody else could replicate.
Re:Strongly disagree (Score:5, Insightful)
Yes because 20+ years was wasted going in the wrong direction and the drugs they made gave false hope instead of working on other things that might have actually worked. No matter how you look at it, they set research back 20 years which is directly 20 additional unnecessary years of horrible deaths from Alzheimer's in the future.
Re: (Score:2)
What drugs? That was one research line .. AB*56. It didn't set research back much. You act like everyone dropped everything and started looking at AB*56 for 20 years. That's not what happened at all.
Re: (Score:2)
Interesting and useful results get replicated. There are LOTS of papers showing that amyloid beta is associated with Alzheimer's. This is a paper looking at a very specific variation, in a mouse model. It is not responsible for pharma's focus on amyloid beta therapies.
Pharma isn't dumb either. They absolutely replicate results before they pour billions of dollars into something.
Re: (Score:1)
I was told by several climate scientologists that there are no fake papers.
Re: (Score:2)
Perhaps he's really commenting that you can't trust Scientologists?
Re:Science working as it should (Score:5, Insightful)
A mistake of this magnitude is unforgivable to the scientists who had knowledge of it.
Doctoring images does not happen by mistake, it's a deliberate attempt to mislead.
Re: (Score:3)
Re: Science working as it should (Score:3)
Sorry, but no. This was supposedly a landmark paper, they not only denied but fought tooth and nail and ruined careers of those that blew the whistle. Moreover this was the source for millions of dollars in research because replication studies are difficult and almost never done because people trust the source.
Doctoring images is an active act of subversion. Itâ(TM)s not a mistake you just make, graphs and images donâ(TM)t doctor themselves.
Re: (Score:2)
The first is reprehensible, and deserves severe punishment.
The doctoring of images...I'd need to know the context (which I probably wouldn't understand). I can imagine that being done just to make a point clearer. Remember, sometimes cartoons are used to render an explanation intelligible.
OTOH, in many contexts, the doctoring of images is just as bad as you imply. I think the decision needs to be left to experts in the field, which I definitely am not.
Re: (Score:2)
That is why you have graphing programs, GraphPad, SPSS, SAS, R, ... they can all render graphs and annotations without distorting the data. In this case they were using Adobe products for "graphing".
Re: (Score:1)
I am a little concerned that you call these "mistakes"? the images were faked, not a case of "ohh, i grabbed the wrong one". This was outright lying. And the author is acting "appropriately" AFTER being called out on it- 18 years later!
18 years of silence
18 years of accolades and prestige from the paper
18 years of fucking up all the research that was based off of these findings- time, money, resources, progress wasted
potentially delayed the actual solution that could impact the quality of life and life te
DNA & photo analysis (Score:4)
Demographics (Score:2)
Doctored images or data? (Score:2)
Unless we're talking images of data.
Re: (Score:2)
Graphs are visual representation of data. There's no difference.
Re: (Score:2)
Images is a bit vague. It's not clear that it's referring to graphs rather than to photos of cells...and those are ALWAYS doctored. They have to be. But you need to explain just what doctoring was done. (Stains, filters, etc. etc.)
Now clearly this was worse than the trivial case, but it's not clear (from the summary) how much worse. Did they substitute one image for another or what.
OTOH, there are good grounds for disbelieving that amyloid bodies cause Alzheimer's independent of this paper. (But good
Re: (Score:2)
Yes you make a good point that we don't know what the doctoring was but since everyone but the guy who did it agreed to pull the paper I'm going to make a reasonable assumption it was pretty serious and not just a non-standard color choice on a cell stain or something trivial.
It had to be fundamental to the point of the paper or no one would care this much.
Agree? Disagree?
Oh, Alzheimer's (Score:2)
I thought at first this was the paper with the AI-generated giant rat phallus.
Re: (Score:2)
That one is legit.
What would the correction have been? (Score:5, Interesting)
I'm troubled that Nature chose to refuse to publish a correction, assuming there was a reasonable correction that could have been made (e.g. substituting the correct images from a previous draft of the paper). Without full transparency, we'll never really know if the doctoring made a material difference in the results.
Re: (Score:2)
I'm troubled that Nature chose to refuse to publish a correction,
The authors have proven unreliable, I'm not surprised they'd be suspicious of giving them a second crack at it with a new set of potentially doctored images. No, the only sensible solution is to flag the paper as fraudulent and move on.
Re: (Score:2)
Since their results have been proved to be unreplicatable it is reasonable to assume there is no reasonable correction that could have been made and still maintain the premise they put forth about what causes Alzheimer's.
A new graph would have just delayed everyone's understanding the paper was crap. All the people with names on it but one agreed to withdraw.
The guy who faked the graph was the only one putting up a fight about it.
Re: (Score:2)
If you want transparency, best get it from people who aren't frauds. And nothing is stopping them from publishing a correction, various websites let you publish anything you like.
Quite meaningless (Score:2)
for which I as the senior and corresponding author take ultimate responsibility
So what is the "ultimate responsibility", are you giving back the grant money to people who would have made better use of them?
Re: (Score:3)
Of course not but he feels really bad about it.
Peer review doesn't work (Score:2)
In general I think most scientists are trying to do the right thing and do publish honest papers but we have seen way too many papers go through and have real world impact that got through peer review and every other filter in place and took years to get pulled. Who knows how many other major results across what fields have time bombs waiting to go off?
Being wrong is totally ok. That's science. But faking results should be a serious crime on the same scale as financial fraud with a possibility of fines a
Re:Peer review doesn't work (Score:4)
In general I think most scientists are trying to do the right thing and do publish honest papers but we have seen way too many papers go through and have real world impact that got through peer review and every other filter in place and took years to get pulled. Who knows how many other major results across what fields have time bombs waiting to go off?
Being wrong is totally ok. That's science. But faking results should be a serious crime on the same scale as financial fraud with a possibility of fines and jail for a conviction. It is definitely a fraud. Having to maybe retract a paper 20 years later, if caught, doesn't seem like enough disincentive to not fake it.
It is awkward for certain. Especially because now every wrong path might have to prove that they weren't practicing fraud. Stanley Prusiner's work on prions was wildly attacked, and the word fraudulent was used a lot. Many claimed that prions didn't even exist. And here were are, he was not only correct, but won a Nobel prize in 1997 for that work.
So yes, a good chance that prions are involved in Alzheimer's https://www.ucsf.edu/news/2019... [ucsf.edu]
The problem is, the retracted work focused on the wrong one, and tried to make it fit instead of acknowledging it didn't.
Now for my own thoughts on the matter, the research in toto is on the wrong side. Prion infection is simply too late. Similar prion diseases like scrapie or mad cow disease (moo, god dammit!) aren't curable, and my money is on Alzheimer's not being curable, or likely even stoppable once it gets into the brain.
So how are the prions getting into the noggin? And if we find that route of infection, how do we prevent it? Is this an inevitable hard stop for aging, or can there be a vaccine if there is an infection route?
And there is the crux. If a lot of people contract Alzheimer's with no actual cure, that doesn't make money for the drug companies. And that is the bad part of this paper, sending research off in the wrong direction. So we get proof that scientist are as human as politicians. Even with messes like this, my money is on them being hella more ethical
Re: (Score:2)
If you have an infection model for Alzhemier's then you could find the cause by transmitting the disease to test subjects (preferably rats or something). If it's prions than it should be able to "infect" tissue samples.
Re: (Score:2)
If Alzheimer's *is* a prion disease, then you don't need a source of infection, merely a protein that has a certain probability of folding the wrong way, and a path of replication.
This *would* align with it only being likely to show up in the aged. And I can think of no obvious reason why it's impossible. But there are reasons for believing that it's not the amyloid bodies that cause Alzheimer's.
Re: (Score:2)
If Alzheimer's *is* a prion disease, then you don't need a source of infection, merely a protein that has a certain probability of folding the wrong way, and a path of replication.
This *would* align with it only being likely to show up in the aged. And I can think of no obvious reason why it's impossible. But there are reasons for believing that it's not the amyloid bodies that cause Alzheimer's.
Certainly the other prion diseases are transmissible. Creutzfeldt–Jakob and spongiform Bovine are spread by eating certain parts of infected animals.
Prusiner et al speaks of a double 'whammy" of prion infection AB with anther called Tau. https://www.ucsf.edu/news/2019... [ucsf.edu] AB*56 , the amyloid protein version of Ashe's paper, was looking in the right direction, but apparently the wrong sub area.
Re: (Score:2)
Transmissible doesn't mean frequently transmitted, or usually encountered because of transmission. The one's we've noticed have been transmitted, but that's partially because of the way we've noticed them. It's not required for a prion disease. All you need is for a protein to have a certain probability of misfolding, and for the misfolded one to tend to create IN SOME MANNER additional misfolded proteins. But this would be selected against by evolution if it happened quickly enough to curtail the repro
Re: (Score:2)
Transmissible doesn't mean frequently transmitted, or usually encountered because of transmission. The one's we've noticed have been transmitted, but that's partially because of the way we've noticed them. It's not required for a prion disease. All you need is for a protein to have a certain probability of misfolding, and for the misfolded one to tend to create IN SOME MANNER additional misfolded proteins. But this would be selected against by evolution if it happened quickly enough to curtail the reproductive period.
As for what causes Alzheimer's...the jury is asking for additional evidence.
My point is not that Alzheimer's is or is not transmissible between humans - the standard path would have to include cannibalism. So it may or may not even be an infection path to Alzheimers.
But to my original post, if it isn't an infectious sort of disease as I suggested but in no way endorsed, and your point of malfolded proteins just happening, then we've pretty well hit a wall with regards to treating Alzheimer's, and it is an incurable disease.
The one drug that slows the progression a little, but
Re: (Score:2)
Well, I'm not sure that would mean we've hit a wall, but it would mean that the treatment would need to jam the reproduction of the misfolded protein.
OTOH, since the actual cause is unknown, it may well be something else. But *IF* it's a prion, it probably isn't the amyloid. (Treatments that really decrease the amyloid levels don't seem to yield improved outcomes. At least that what I understand Derek Lowe to have said.)
Hot Take (Score:2)
Hold public executions for all those researchers.
The high number of fake papers (Score:2)
Free speech be damned, science publications are NOT The Onion and the consequences of fraud can be severe. This has reportedly included the sale of dangerous medication, but actual examples are less important than the potential.
Science speech should be placed in a special legal category where both the presentation and publication of fraudulent articles carries penalties, including the automatic rescinding of academic qualifications by scientists indulging in fraud.
There should also be an official rescinding