Your daily negative XMRV study: XMRV is PCR contamination

[kurt]

Kurt,
Based on your responses, I suspect your background is similar to mine -- the physical rather than the biomedical sciences. In a former life, I both wrote and reviewed scientific papers in my tiny little corner of the research world. In the physical sciences, scientists are indeed highly dedicated to the scientific method, and to honest, accurate science. You can rely on scientists working in good faith.

I have been told frequently in the past couple of years that this is not necessarily true in biomedical sciences. What I've seen in the past year has led me to believe this is true. I suspect it stems from an inherent lack of understanding of mathematics and statistics, which leads to a much lower standard of rigor than physical scientists expect. This allows for what I consider to be an unconscionable amount of bias and intrusion of opinion into what ought to be pure science.

You are close, I worked in an interdisciplinary field as a systems scientist. I also worked in applied cognitive science (military training), and some coursework background in the physical sciences (physics and chemistry), so you are partly right. Maybe I am too trusting and I do know what you mean about problems with misapplication of statistics in the biological sciences, have had to battle that in my own field. But this is always a two-way street. I have no implicit faith or trust in WPI either, no more than any other lab. They have produced a study that is either incorrect or they did not accurately describe their process, because these other researchers should have confirmed their results many times over by now. Finding a virus should not be so difficult, if it is really there.

Many CFS patients seem have adopted the idea that if a lab confirms XMRV then they are conducting good science, and if they are not able to confirm XMRV then they are shoddy scientists, bad at math, or have a hidden agenda. I don't know where this idea comes from, but just is not reality. The evidence of these studies will all have to be weighed together and tests validated in blinded studies before it is even possible to figure out who is doing good science.

 

[SOC]

I worked in an interdisciplinary field as a systems scientist. I also worked in applied cognitive science (military training), and some coursework background in the physical sciences (physics and chemistry), so you are partly right. Maybe I am a little too trusting and I do know what you mean about problems with misapplication of statistics in the biological sciences, have had to battle that in my own field. But this is always a two-way street. I have no implicit faith or trust in WPI either, no more than any other lab. They have produced a study that is either incorrect or they did not accurately describe their process, because these other researchers should have confirmed their results many times over by now.

It certainly does go both ways. I am more confident of the Lombardi, and Alter/Lo papers than I am of others because they received extremely rigorous reviews by highly reputable, top-ranked journals. When papers of equal quality come out for the psychological side or lack-of-XMRV side, I'll give them equal credibility.

I am not 100% convinced that XMRV is the cause of ME/CFS. It's still not scientifically proven. The evidence is very good at the present time, however, so I consider it the most likely theory and I want to see a lot more research done in that area. I also want to see research in other likely areas -- which do not include psychological theories.

I won't argue that scientists should have been trying to replicate, rather than trying different methods. As to why they haven't, I'm not in a position to say.

 

[omerbasket]

They almost don't mention the serology findings by numerous studies - and especially in the WPI's studies, and therefore their paper is a little bit ridicilous - if it's a PCR contamination, why does it leave behind antibodies?
Ofcourse the antibodies could be of another human gammaretrovirus, but that would be an unlikely coincidence, and anyway - does it really matter if it's one gammaretrovirus or another? Not very much, especially not when you see the other possibility: Stopping the search for gammaretroviruses in the human population, and potentially, let millions of people ruin their life, or die, because of a gammaretrovirus infection.

 

[omerbasket]

You are close, I worked in an interdisciplinary field as a systems scientist. I also worked in applied cognitive science (military training), and some coursework background in the physical sciences (physics and chemistry), so you are partly right. Maybe I am too trusting and I do know what you mean about problems with misapplication of statistics in the biological sciences, have had to battle that in my own field. But this is always a two-way street. I have no implicit faith or trust in WPI either, no more than any other lab. They have produced a study that is either incorrect or they did not accurately describe their process, because these other researchers should have confirmed their results many times over by now. Finding a virus should not be so difficult, if it is really there.

"SHOULD not be so difficult"? The virus doesn't listen to you. And that is the exact arrogance that I'm talking about. We don't know! We just don't know if something is causing us not to find the virus even if it's there. That is why replication studies, of numerous methods that were successful at finding the virus, along side with testing the same kind of cohort, are needed.

You know, if it's a contamination it SHOULDN'T have been found so many times in the WPI's/NCI's/CC's studies, and especially not when it's also the studies of Lo/Alter, Hanson, Singh, Silverman, Danielson etc. etc. etc. Especially not when the studies include serologies.

Those who doesn't do a replication study, doesn't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distruct us from the truth if they don't find the virus and if the primer study was, infact, correct.

Studies results can be true and correct even if they are not validated. They just need to be verified, and there is a big difference. Until today, no one, and that means not a single researcher, tried to verify WPI's findings. Even though, a few could actually validate them, or their major finding (Lo/Alter, Hanson and De-Meirleir, and regarding the finding of XMRV in the normal population, also the Japanese and one can assume that also the respiratory tract secretions' study).

 

[acer2000]

I think some of these studies are displaying a bit of circular reasoning. Its entirely possible that XMRV both infects humans and causes illness, and can be found in cancer cell lines derived from humans that can be used in lab experiments. In fact, if XMRV really causes prostate cancer, I'd almost expect that some PC cell lines have XMRV in them because thats probably what caused the malignant transformation in them in the first place.

Its not an either or scenario...

Plus, for contamination to be the issue, you really have to explain the fact that the samples were blinded, yet the only ones that came up positive for the pathogen happened to be the ones from the patients with symptoms. Its really hard to selectively contaminate only samples from sick people.

 

[Cort]

"SHOULD not be so difficult"? The virus doesn't listen to you. And that is the exact arrogance that I'm talking about. We don't know! We just don't know if something is causing us not to find the virus even if it's there. That is why replication studies, of numerous methods that were successful at finding the virus, along side with testing the same kind of cohort, are needed.

You know, if it's a contamination it SHOULDN'T have been found so many times in the WPI's/NCI's/CC's studies, and especially not when it's also the studies of Lo/Alter, Hanson, Singh, Silverman, Danielson etc. etc. etc. Especially not when the studies include serologies.

Those who doesn't do a replication study, doesn't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distruct us from the truth if they don't find the virus and if the primer study was, infact, correct.

Studies results can be true and correct even if they are not validated. They just need to be verified, and there is a big difference. Until today, no one, and that means not a single researcher, tried to verify WPI's findings. Even though, a few could actually validate them, or their major finding (Lo/Alter, Hanson and De-Meirleir, and regarding the finding of XMRV in the normal population, also the Japanese and one can assume that also the respiratory tract secretions' study).

I have come to the conclusion that it really shouldn't be that hard to find. Why? Because it wasn't hard for the WPI to find. They didn't use particularly sensitive PCR. They went off the very clone they argue should not be used now - the VP62 clone. They didn't use a particularly large amount of DNA (if I got his right) to search for the pathogen...yet they found it in almost 70% of the study participants...Small lab using not particularly sensitive equipment searching for it in the only they knew then but is primitive now (VP62) and readily finding it - I think it really shouldn't be that hard to find.

I was swayed by the idea that you need to culture the virus but at some point I went back to the original paper - and there it was in black and white - XMRV showed up in spades using their kind of normal PCR equipment....using VP62.....blah blah.. Now they have better antibody tests, they know other sequences to search for (one of the sequences WPI looked at turned out be very variable if I remember correctly) - they should be able to find it more easily now actually and they still cannot find it....That's just my laymen's guess.

We'll really know when the BWG studies come out. I guess that will be awhile unfortunately but I think they will either make or break XMRV....They could turn everything around in a heartbeat or really, really (really) put a damper on things.

 

[Mark]

Bear in mind: As soon as the original study came out, Wessely and Reeves were straight in there, they told us this a year and a half ago: Our bet is it won't be replicated. I wonder how they guessed/predicted that?

It then took Wessely about 2 months to get a study finished that didn't find anything. Others have been conducting partial studies using different techniques, that don't find anything at all, ever since - it's become quite fashionable...

Omerbasket makes some good points...I've corrected a few typos...

You know, if it's contamination it SHOULDN'T have been found so many times in the WPI's/NCI's/CC's studies, and especially not when it's also the studies of Lo/Alter, Hanson, Singh, Silverman, Danielson etc. etc. etc. Especially not when the studies include serologies.

Those who don't do a replication study, don't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distract us from the truth if they don't find the virus and if the original study was, in fact, correct.

Studies' results can be true and correct even if they are not validated. They just need to be verified, and there is a big difference. Until today, no one, and that means not a single researcher, tried to verify WPI's findings. Even though, a few could actually validate them, or their major finding (Lo/Alter, Hanson and De-Meirleir, and regarding the finding of XMRV in the normal population, also the Japanese and one can assume that also the respiratory tract secretions' study).

I think I'll repeat this bit again, because I rather think if Gerwyn were here we'd have heard it said more often...

Those who don't do a replication study, don't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distract us from the truth if they don't find the virus and if the original study was, in fact, correct.


Bottom line: this won't be over until the WPI (and the PC researchers) say it's over.

 

[ixchelkali]

Greg Towers lost all credibility with the comments he made following his last paper: "Our conclusion is quite simple: XMRV is not the cause of chronic fatigue syndrome." That clearly overstated the evidence of their study.

So now I'm looking at a study by Greg Towers at University College London, published in Retrovirology (where no negative XMRV study is too shoddy to go unpublished), compared to a study by Bob Silverman at the Cleveland Clinic, published in the Proceedings of the National Academy of Science (PNAS). Which am I going to give more credence to?

It's not that I think that if someone publishes a negative study, it must be shoddy work and they must be part of a conspiracy against CFS patients. I am in favor of rigorously conducted scientific studies, regardless of the conclusions. That's how you arrive at the truth, and that's what we need.

I do, however, think that Greg Towers has done shoddy work and has engaged in a smear compaign to discredit XMRV researchers. What he does perverts scientific inquiry and clouds the truth. That does a disservice not only to ME/CFS patients and to the retrovirologists whose reputations he besmirches, but to all scientists who do carefully conducted work.

 

[SilverbladeTE]

Kurt
there's been increasing numbers of articles and complaints by folk within the system, that regarding medicla research, especially pharma' reserach, it is UTTERLY broken, because of corruption and ludicrously bad science (like gross misunderstandings of how statical anyalysis actually works, hell even as a layman I saw some gross errors in a few, which later got picked up on. Alas, unlike say physics, where the field is unsruprisingly innately rich in the understanding of mathematics/statistics, medical research does not seem so nearly competant)

IIRC was an ex editor of the new England journal of medicine who had retired in disgust as a majority of published items had been falisfied, "adjusted" or whatever.
simple starting point to explore this kind of topic
http://www.naturalnews.com/bribes.html
the corruption is IMMENSE.

New work shows the Research & Developement costs for pahrma corps are even lower than the 1/3rd previously known, pharma corps had prior claimed R&D cost 2/3rds+ of their budgets, which was a lie, now it's more like 2/6ths gross profit, 3/6ths "advertizing" (which includes "bribery") and only 1/6th on R&D!

bah whole systems rotten to hell.

The reason this is, is because when science collides with profit and power, corruption is INEVITABLE.
Governments are scared sh*tless of a major health crisis, like AIDS, especially because HIV/AIDS was tainted by vile practices/corruption and that would trigger public anger would threaten public order and whether politicians got re-elected or not (and hence why were are being so abused, as if truth came out, there'd be hell to pay).

Government takes HUGE bribes from pharma corps, "campaign funding" is bribery folks, like it or not, it's undue influence. IIRC, 2003, UK Labour party conference was part funded by UNUMProvident insurance company as an eye opener on how widespread this threat to democracy is.
USA's supreme court isnanely took off any limits ot the donations politicians could recieve or form who, so imagine say a US senator geting $10million from a pharma corp or China...?

And so on and so forth.

Even in academia, science gets corrupted by ego, alas it's jsut Human Nature, but it's usually discovered because that system isn't so rank driven by profit and silenced by horrendous contracts
after scandals in 1980s were exposed by whistleblowers, pharma corps started to bring in "antiWhistleblower contracts", as did some other companies
while claiming these were to stop industrial espionage, they were really to silence anyone from disclosing criminal or immoral actions by pharma corps. If anyone talked, they were open to huge civil fines


And guess what section of the medicla world seems to be the most heavily supplied with pharma corp "advertizing" etc?
Psychiatry. See US investigations into that.

 

[free at last]

I think Corts points need a answer before i stop worrying now. Hes been saying it for some time, and its the one thing that has bothered me the most. Because how can a pcr study that is more sensetive find less ? worse than that infact, how can many pcr studys that are on par with WPIs pcr testing, keep failing so consitently ? yet VIPDX can seem to find it easy.

if all the studys were sent to vidpx, most likely by now we would have a host of positive studys, and im not talking culture or serology here im talking figures from VIDPX with JUST PCR. Somethings not making sense. how ever much it pains me to think it. something is just not adding up. and if someone can respond to Corts very valid observation, and put my ( corts ) and everybody elses mind at rest on this question please do so, because i actually want to belive this is still rolling along nicely.

Its just it doesnt actually feel that way. Either there really is deep conspirecy of the likes even i have trouble accepting. or the PCR evidence is somehow being affected between what the WPI got, and what virtually everybody else seems to be getting. so what is the problem thats preventing researchers detecting XMRV the way the WPI did in science ?

I understand the culture, serology and healthy control versus patient positive detection evidence ( blinded ) do seem to suggest yes its a virus they are finding in CFS patients. but none the less. the questions put foward by Cort, are being observed by other researchers, and there conclusions are being drawn from.
rightly or wrongly.

This problem seems to me to need a answer, as yet i havent seen one that makes me relax. if anyone can, im listening intently right now. I in no way would ever attack the wpi or there work. as infact they have been the greatest hope in my life, for answers, for understanding why im locked away in the winter for fear of returning to that hell i lived. for the way my symptoms improve for a while then relapse, even after weeks of fairly normal health. i want them to be right, i so do.

 

[eric_s]

I can't make sense of it neither, Free at Last.

And the crazy thing is that on both sides it's as if people are going "all in" in poker.

What makes me believe in XMRV is that in the positive studies the percentages seem to be more or less the same for healthy controls (1 digit), prostate cancer (around 25%) and ME/CFS (2/3 or higher). And that we have very credible people like Ruscetti (HTLV discovery), Alter (Hep C), Silverman and others that have not backed off until now.

I get some comfort from the thought that even besides XMRV we have the CSF findings and also the findings of Montoya and Lerner who show that antivirals can help. So there are definitely leads of which i'm quite convinced they can lead us further.

 

[bullybeef]

http://www.retrovirology.com/content/8/1/13/comments#471696

 

[lansbergen]

I think I'll repeat this bit again, because I rather think if Gerwyn were here we'd have heard it said more often...

Those who don't do a replication study, don't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distract us from the truth if they don't find the virus and if the original study was, in fact, correct.


Bottom line: this won't be over until the WPI (and the PC researchers) say it's over.

Gerwyn posted a comment.

 

[kurt]

Bear in mind: As soon as the original study came out, Wessely and Reeves were straight in there, they told us this a year and a half ago: Our bet is it won't be replicated. I wonder how they guessed/predicted that?

The opinions of Wessely and Reeves have nothing to do with whether or not XMRV is a valid scientific finding in CFS. There is no mystery behind their opinions, they believe in other explanations for CFS.

It then took Wessely about 2 months to get a study finished that didn't find anything. Others have been conducting partial studies using different techniques, that don't find anything at all, ever since - it's become quite fashionable...

..
I think I'll repeat this bit again, because I rather think if Gerwyn were here we'd have heard it said more often...

Those who don't do a replication study, don't find XMRV and run to the press to publish it - they are conducting bad science. That is because at least at the current state, good science demands that in order to publish a negative study, this study should be a replication study - because other studies can only distract us from the truth if they don't find the virus and if the original study was, in fact, correct.
Bottom line: this won't be over until the WPI (and the PC researchers) say it's over.

Not finding XMRV, and rushing to press is not an indicator of bad science. It is simply a negative finding. I am surprised that anyone with science training would persist in this type of conspiratorial thinking. There may be a conspiracy against ME/CFS in some government or insurance agencies, but not in retrovirus laboratories. They would like nothing more than to find a new contagious retrovirus. Many of the negative studies come from labs expert in retrovirus detection.

Is this a mystery? Yes. Is it a broad scale epidemic of bad science? Absolutely not.

 

[alex3619]

Hi bullybeef, this is a good comment on the link in post 32, by Gerwyn (the same Gerwyn we know?) I completely agree with it. Bye, Alex

 

[SOC]

Not finding XMRV, and rushing to press is not an indicator of bad science. It is simply a negative finding. I am surprised that anyone with science training would persist in this type of conspiratorial thinking.

I strongly disagree.

Rushed research and rushing to press is a very strong indicator of bad science. I speak as a writer and reviewer of scientific research papers (although in an entirely different field). This is not conspiratorial thinking, it's a reality of the scientific process. Good research needs to be carefully, accurately, and thoroughly done -- see the Lombardi, and Alter/Lo papers. It takes time, often well over a year.

Good published research is carefully and thoroughly reviewed by experts in the field. This often means questions, clarifications, and rewrites to make the paper as complete and accurate as possible. This takes time -- many months as a general rule -- as we saw with the Lombardi, and Alter/Lo papers.

Research that is performed and published in a matter of weeks, even several months, is questionable. It is not necessarily bad, but it requires extremely careful assessment on the part of the reader. It begs the question of how the researchers did careful design of experiments, selection of samples, and accurate performance of experiments in such a short time.

Any scientific paper rushed to publication did not receive the level of review expected in strong papers. Reviewers are supposed to catch the flaws in the paper -- poor experimental design, poor data analysis, poor sample selection, bias in discussion, inappropriate conclusions. Unreviewed and rush-reviewed papers do not receive that kind of vetting and consequently, the reader cannot have the same degree of confidence in the paper that they have in papers that have been thoroughly reviewed.

There may be a conspiracy against ME/CFS in some government or insurance agencies, but not in retrovirus laboratories. They would like nothing more than to find a new contagious retrovirus. Many of the negative studies come from labs expert in retrovirus detection.

Is this a mystery? Yes. Is it a broad scale epidemic of bad science? Absolutely not.

I understand you have a background in science, Kurt, and that is very valuable in evaluating a lot of this research. However, I suggest that you need a more extensive experience in science, and in particular research and publication of science before you can convincingly say,

...rushing to press is not an indicator of bad science. It is simply a negative finding. I am surprised that anyone with science training would persist in this type of conspiratorial thinking.

For the benefit of those worrying about the negative studies -- all research is NOT created equal.

Have any of the negative papers been published in a top-ranked journal like Science or PNAS? Not even close, as far as I recall. They just don't compare. Papers published in lesser journals, especially unreviewed publications, are not necessarily bad, but they are much more open to question and simply don't carry the same weight as the best papers.

I think Dr Mikovitz has said something similar in the past. It's a fact known to all good researchers -- Good research takes time.

Quick and dirty studies rushed to press look a lot more like a desire to influence the media than they look like good science designed to convince responsible researchers in their field.

 

[bullybeef]

Is it not also a marker for bad science if the paper is a paid to print journal: Erlwin et al, Plos One, Jan 2010. Incidentally, to those whom have forgotten or are unaware, their paper took just three days for review. Probably enough time for the cheque to clear!!

Says everything really.

And it is still taken seriously, and still mentioned as one of the main negative XMRV papers.

BB

 

[SOC]

Is it not also a marker for bad science if the paper is a paid to print journal: Erlwin et al, Plos One, Jan 2010. Incidentally, to those whom have forgotten or are unaware, their paper took just three days for review. Probably enough time for the cheque to clear!!

Says everything really.

And it is still taken seriously, and still mentioned as one of the main negative XMRV papers.

BB

As a general rule, yes.

If the paper is in a paid to print journal, then it has less credibility than a paper in a competitive journal. That does not necessarily make it bad research, but the probability is much, much higher.

PlosOne review, for example, is minimal. Last I heard the reviewer is not necessarily even someone directly in the field of the paper and almost certainly not a top researcher in the field. That's how review gets done in 3 days instead of 3 months.

Not everything published in PlosOne is bad research. However, good researchers who publish in PlosOne do so to get the word out quickly, usually for the benefit of other researchers on some relatively small topic, and they understand the credibility gap. They certainly don't publish there to refute primary discovery and confirmation papers published in top journals like Science and PNAS. That would be akin to claiming your Little League team is as good as the winning World Series baseball team because you both won a game. [snicker]

 

[Esther12]

I'm confused by this 'paid to print' thing.

I've seen it claimed elsewhere that most journals charge those publishing. From what I've seen on-line, academics don't seem to respond well to CFS patients making a thing out of certain studies being paid to print, so regardless of it's merit as a point, it might not be politically beneficial to make a thing of it.

 

[Noah_Scape]

I remain open-but-sceptical o the idea of XMRV being "a human pathogen", and possibly the cause of CFS.

Do the authors of the study quoted above have any REASONING to back up the last two sentences?
I.E. - "We propose that the patient-derived sites are the result of PCR contamination. This observation further undermines the notion that XMRV is a genuine human pathogen."