Discussion in 'XMRV Research and Replication Studies' started by Jemal, May 4, 2011.
Bottom line: there's still much confusion
Hi Cort, Thank you for your comments, it brings up an important point. My claim is that nobodycan disprove the contamination claims. We can only show that they are inconsistent with the data. This is why I say it is falsified, which I was hoping was clear does not mean false. Falsification in my argument is an assignment of a false value due to considerable evidence, but it can itself be falsified. It is less about a truth value (true/false) then a function of determining truth, which is always conditional. By saying it is falsified, not false, I hope that people will be be alert to the possibility that even one single additional datum can change everything. This kind of data comes from research. Every new study needs to be examined.
One really good study could change everything. For example, we might find that both the contamination hypotheses and the association hypothesis are wrong. There is a genuine mystery here, and we are all (including professional scientists) really speculating about things based on inadequate data.
I usually don't talk about this stuff on PR as it is highly confusing, even for me.
I agree with your entire post, the only diffeence between our positions is a view of the context. I wrote:
To date WPI have not done that within any standard accepted by their peer group, thus the linkage between XMRV and M.E/CFS is unproved - and that is how it remains until a proof (not a dis-proof) is given.
Popper notwithstanding, the reality is that something akin to 'proof' is now needed for the wider medical research 'peer group', to see an XMRV association with CFS, as anything other than 'artefactual'. The language of falsifiability will no doubt be used in respect of the Lipkin and BWG studies, but some solid arithmetic is probably going to be needed to convince anyone outside of the WPI that an XMRV/CFS association, has substance.
Again I entirely agree, what is interesting though is how the utilitiy of 'Lombardi's Hypothesis' has been lost in the language used by the WPI, something which I would argue has been compounded by the need to support the validity of the WPI licensed commecial test. There isn't just the LH involved here, there is also the hypothesis that 'XMRV can be detected in the blood of people diagnosed with M.E/CFS' - where an hypothesis moves from scientific utility to technological and commercial exploitability, some 'evidential proof' that what is being provided/sold is valid is required - an unfalsified hypothesis is not generally thought sufficient to define whether a product is fit for purpose. Something in the range of a $millon has been spent by individuals on these tests, and falsifying the 'XMRV can be detected in the blood of people diagnosed with M.E/CFS' hypothesis inevitably raises questions about the commercial tests and the context in which they have been marketed, which in one not easily seperated from the research.
I think, given the diversity of the XMRV the CDC found that we can assume that the WPI also knows how diverse this virus can be, even though their published sequences to date are similar to 22rv1 XMRV.
If the CDC's findings are correct, the XMRV they found is not similar enough to the 22rv1 cell lline for any theory to be based on a connection between the two.
I think it highly unlikely, given that the XMRV the CDC found mutates readily, that XMRV can have originated in the way Dr. Coffin suggested. This theory could only convince if XMRV was more stable than it is, as it relies on stability and continuity in the sequences for the theory's logic. With a very variable virus, it is much more likely that any similarities are just coincidence.
This theory always seemed far-fetched to me because all retroviruses mutate readily. They are not able to copy themselves with great fidelity, and they also recombine with each other inside cells.
Just because XMRV can be created from two half sequences, does not mean that it originated in that way.
If the WPI are on the point of publishing new sequences for all the XMRVs they find, surely this theory will become hopelessly irrelevant.
This is an appeal to consensus. I know various 'skeptic' groups (and wikipedia) are slaves to the notion of consensus (misreading Kuhn as authoritatively prescriptive rather than descriptive a lot of the time), but the notion of consensus is highly problematic. You are presuming that some hypothetical and unqualifiable 'medical peer group' has a sound a priori belief in the 'artefactual' (when it's really not sound). And this is the problem. We are in flat earth territory again.
The burden of verification is equal to all parties in science. You are assuming that the WPI group have a greater burden of 'persuasion' than anyone else, based on appeal to another unquantifiable and unqualifiable 'consensus'. The need for 'solid arithmetic' (whatever you actually mean by that) applies no more to the WPI group than any of the other parties.
As for the 'lost in language', surely you must understand that ALL science is in danger of becoming 'lost in language'? When Singh decided to condemn the possible use of antivirals, she was 'losing' her 'science' in language. When McClure used the ludicrous 'you better be 1000% sure' comment she 'lost' her 'science' in the language. When the Wellcome Trust put out their premature assertions in a press release, they lost their science in the language. The PACE authors have always lost any science in language (because of psychiatry's tendency towards the Freudian reliance on ad hoc literary constructions, for just one reason). But their rhetorical attempts to discredit their critics has again, lost any hope they had of 'science' in language.
This is why 'the scientists' should be a lot more careful in what they claim. Those attempting to claim the WPI's findings as false have not been careful enough in what they are claiming, likely because of other reasons apart from 'science'. The attacks on the ME community for their daring to engage with the science have also contributed to the loss of science into ideology (your 'language').
Hi IVI, if you are saying that with this much evidence against the WPI has a need to provide something substantial to justify its claims, then we are indeed in agreement. There have been so many negative comments, with so much circumstantial evidence against them, that it makes it hard to give credence to their claims if you don't take the time to examine everything carefully. Most researchers who are not in the field, most medical practitioners, and most in government will not take the time to consider everything. This is as much about the politics of science and government as it is the science itself.
If the BWG and Lipkin studies have sufficient power, that might convince many, whatever the result. If they are not sufficiently conclusive, or for some reason have major problems we have not forseen yet, then there will still be debate. I still think XMRV is a good hypothesis, but many retroviruses, and potentially many viruses, may be at least as good a fit if you discount the Lombardi paper in Science. There is even growing evidence that the immune deficits we frequently have are at least in part due to genetics - which opens the door for a wide range of viruses or other pathogens to be co-causal with genetics.
As I see it this is a mystery, and we need definitive answers. Ultimately though, either Lombardi et. al. and associated scientists are right, or they are wrong. If Lipkin is correct that he might have results before the end of the year, we might finally have the answers we need so that we can either move on to proving causation, or re-emphasize other lines of research. I am wondering if the BWG is going to present some results at the IACFSME conference in Canada later this year. It would be nice if we heard some more of the science at the 20 May conference too.
PS just to be clear my preceding comments in this post are based mainly on the politics. Science is not devoid of politics unfortunately. The science behind Lombardi et al is strong enough to warrant a good blinded controlled test like BWG or Lipkin. The politics is used to convince everyone to maintain the status quo, even here in Australia where our own health minister dismisses XMRV automatically. To break through the extraordinary barrier to taking ME/CFS seriously, we need extraordinary evidence.
Yes, I agree that this is a problem... The WPI have now issued a statement that new sequences have now been deposited with genbank: "Multiple sequences from these three thousand samples have been submitted to GenBank and are awaiting publication."
Yes, I agree that there may have been a certain amount of shifting, in relation to the methodology, since the original Science paper. But this is understandable, as they have now had about 2 years to further develop and enhance their knowledge about the subject, since they carried out the Science paper research.
The point I was making is that they haven't changed their position on their fundamental conclusions in the Science paper.
Many scientific commentators have had to fundamentally change their positions, due to the unfolding research developments.
Yes, Singh's results need to be considered seriously, but the main point that I was making is that they were not conclusive in any way. If the Singh results help the WPI get a better understanding of their own research, then obviously that will be very useful. I think we all need to keep in mind that XMRV research is still very much a work in progress.
I don't agree with this because I believe that a zero result just tells us that the methodology was not adequate to detect XMRV, or that XMRV was not there. It does not tell us that XMRV was not there.
Yes, quite possibly. But the genetic variability of their sequences, suggest that this is not an issue.
I might be wrong about this but I understand that it is possible to process non-blood negative controls (i.e. definitely negative samples) through the entire processing routine, as if it was a blood sample. Then you can test this negative control sample for XMRV contamination. But I don't have much insight into this process.
My main point is that the WPI's work is still valid, and I believe that the evidence seems to be continuously moving towards the WPI's interpretation of the research.
For example, the high profile paper that stated that XMRV was only a product of a cell line, has been disproved by the new CDC paper.
There are many other examples of how the establishment view of XMRV has shifted, as I explained in my earlier post. (I seem to have used some bad wording in my examples, so I've gone back and changed my wording slightly - see my next post.)
Some of my earlier post seems to have been slightly misinterpreted, because I wrote it badly, so I've rewritten some of it:
It's interesting to note that all of the other previous claims against the WPI's work have now been rebutted by continuing research developments:
- First of all, critics of the WPI said that XMRV was not a real virus - that XMRV was just a false positive reading caused by some other product. But that has been disproved. XMRV is a real virus.
- Then the WPI's critics said that XMRV was purely mouse contamination from mouse DNA. But that has been disproved. XMRV is a real virus.
- Then the WPI's critics said that the XMRV wasn't a real human virus, but was a mouse virus. But everyone now agrees that XMRV is a real human virus.
- Then the WPI's critics said that the XMRV that the WPI detected was purely the product of a cell line, and not a wild virus. The WPI has now deposited a wider variety of strains to genbank, awaiting approval. Even the CDC has now confirmed that XMRV is a real wild human virus, and Switzer has detected 3 entirely new strains of XMRV.
- Then the WPI's critics said that the established testing methodologies, used in the 0/0 studies, were totally adequate to detect XMRV in the blood. But now the CDC says they cannot detect XMRV in the blood of XMRV-positive prostate cancer patients, using established technologies.
(The latest significant, but widely overlooked, CDC study by Switzer et al., demonstrates that XMRV is a human virus, and that it is not detectable in the blood by established technology or procedures, even in XMRV-positive prostate cancer patients.)
- Now the only thing left for the WPI's critics to claim seems to be that XMRV is 'definitely' and 'absolutely' not associated with CFS or ME. I'm now waiting for this to be disproved.
Depends what you mean by 'extraordinary'?
I would question your last point (bolded by me here), if you mean there should be a greater burden of evidence on certain groups to show ME/CFS needs to be taken seriously (although I don't think this what you're arguing).
Psychogenic explanations for 'CFS/ME' have never been safe, logically or scientifically, but have relied upon certain power structures to manipulate people to 'believe', in a quasi-religious manner, in them, to accept them without rigorous interrogation. The same sort of problem happened with demonic possession (or godly possession) explanations for epilepsy (the so-called 'holy disease') and other natural phenomena in history.
This problem permeates XMRV, or any other research that attempts to establish the abnormal organic processes that appear to be going on in at least some people given ME or CFS diagnoses. In a way, it's like saying you need extraordinary evidence to demonstrate to a Faith healer and his congregation it wasn't him and god who cured the kid with meningitis, it was the antibiotics (we're talking big time retreat to committment when we talk psychogenic explanations, unfortunately). You might never convince the faith healer and his congregation (evidence of antibiotic cure successes might never be 'extraordinary' enough), so therefore you should never give antibiotics.
It is possible that the 'extraordinary evidence' that is needed is to actually show how unsafe psychogenic explanations are and disseminate them widely (I, for example, believe that is the case). In which case it is only 'extraordinary' in that few people have realised it before. As far as XMRV is concerned, whatever the politically motivated misreprentations of the issue, the Lombardi findings have not (yet been and may never be) 'disproven'. That in itself might be described as 'extraordinary'.
But the problem is that a lack of scientific rigour permeates the comments about XMRV from those claiming scientific authority and expertise. So we're left with a position that many commentators (particulary the so-called scientists) out there are acting like the faith-healing congregation, or the faith healer! Evidence, no matter how 'extraordinary' or rigourous, will not remove them from their retreat to commitment (belief in psychogenic explanations for CFS). I don't think we (the ME community ,whether sufferer or supporter) need to accept that impossible situation though, because to do so (to accept that somehow certain evidence needs to be more 'extraordinary' than the evidence of other parties) is to acquiese to unreason.
I say all this as someone who has no position on XMRV yet, but who wants to see good quality science prevail in order to try and secure best possible outcomes for her daughter, and is seeing precious little evidence of good quality science in the 'anti-XMRV' position so far.
This is a good summary of the situation, as far as I can judge. It could get lost in this thread. Shame to waste it.
Have you thought of blogging it?
I'd like to add that this indirectly supports the findings of the animal model studies as well. In both the monkey study and the mouse study, they found that the virus was difficult to find in the blood - despite being in the tissue.
Until someone actually shows that X contaminated the WPI then of course they cannot prove the study was wrong. This is the weight of evidence building up - not a smoking gun - 26 studies that have not found it....a very comprehensive study that looked for XMRV 9 ways that was done in collaboration with the WPI....the possibility that XMRV is a lab creation, the lack of genetic diversity (thus far )...it's the weight of evidence....that is driving conclusions by these researchers.
I am waiting until the BWG and Lipkin studies are in.
But any scientists well-versed in the scientific method would- should- be aware that it's not quantity that counts, it's validity. If constant publication of poor quality, non-replicative studies is 'driving conclusions by these researchers', than the researchers don't know their scientific method, hence the frustration of, ironically, many of a much-despised lay patient community, who, as it turns out, DO know about the scientific method! (Who'd have thunk it? )
I agree Angela - I have been so impresed by the depth of knowledge of some patients, awesome - they put the scientists to shame with their so called replication nearly/not quite studies.
But Cort, the scientific knowledge and understanding of XMRV is developing very very slowly. It is only in its infancy right now. So I don't believe that anyone should be drawing conclusions from negative results yet, unless they absolutely disprove the WPI's research.
I know that I keep talking about the new CDC study, that's been widely overlooked, but in my opinion, this was quite a significant study...
The CDC have now confirmed that XMRV is a real wild human virus. Not just a human virus, but a wild human virus that patients are actually infected with.
The CDC's study disproves the theory that XMRV is only the product of a cell line, and does not exist in the wild. It confirms that XMRV has actually infected humans in the wild. This in itself doesn't tell us anything directly about the WPI's research, but it does confirm that the WPI's conclusions have the possibility of being correct, especially as the WPI has now confirmed that it has now detected a wide variety of MLV-related viruses, some of which are now awaiting approval for genbank.
The CDC has also shown that they are unable to detect XMRV in the blood, in XMRV-positive prostate cancer patients, with their current established technologies, due to low copy numbers. This demonstrates and confirms that XMRV exists in extremely low copy numbers in the blood. Doesn't this demonstrate to us how difficult XMRV is to detect, even by the experts? If the CDC had tested only the blood of the cancer patients, and not the tissue, then it would have been another 0/0 study, and they would have concluded that XMRV does not exist in prostate cancer. As it is, they have concluded that there is no association between XMRV and prostate cancer, but this seems a premature conclusion to make seeing as they are having so much trouble detecting XMRV, and other studies have found it in such a higher proportion of patients.
Yes, there is a possibility that the WPI are wrong, and yes there is a possibility that Singh's study came to an accurate conclusion.
But the Singh study was not definitive or conclusive in terms of XMRV research.
As Lo said in his video presentation, there only needs to be one very small (and seemingly insignificant) difference in methodology and that can make all the difference between detection and non-detection, especially with such an unknown virus with titres at the limits of detection.
I agree that, on the face of it, things don't look great for the XMRV-ME association, but even if we had 100 negative studies, and only 1 positive study, that does not mean that the 100 studies are accurate and the 1 study is inaccurate.
As we've agreed, we have to wait for the science play out before we can make any firm conclusions.
That is not what Alex's argument amounted to. Alex amplified points about falsifiability of hypothesis - not about whether 'findings' are provable. What Alex defined as the Lombardi Hypothesis was tht which is implied in the introduction to the science paper: The recent discovery of a gammaretrovirus, xenotropic murine leukemia virusrelated virus (XMRV), in the tumor tissue of a subset of prostate cancer patients prompted us to test whether XMRV might be associated with CFS.
When it comes to real world phenomena - then evidence is required and 'proof' of evidence has to be demonstrated if others are to accept the reality of the claimed phenomena; in science the requirement (the source of 'proof') is that the results of experiment should be reproducable. To argue that only by an absolute strict adherence to a single protocol can a result be reproduced requires an extraordinary explanation of why a result can only be achieved by that strict adherence. In the absence of a sustainable argument for the need for a wholly undeviating replication of Lombardi et al, then simply saying that all those studies which fail to reproduce the Lombardi results are 'bad science' is absurd. It maybe that the Lipkin and BWG studies will demonstrate that XMRV, as a non contaminant, can be detected in blood, and is indeed present in some people with a diagnosis of M.E/CFS - and that will indeed provide a 'proof' of the phenomenon that Lombardi et al claimed to have observed. Conversley even if neither the Lipkin nor the BWG studies find XMRV in the blood of people with an M.E/CFS diagnosis, then the Lombardi Hypothesis will remain unfalsified because that hypothesis (as defined by Alex) is not reliant on XMRV being identified in the blood of M.E/CFS diagnosed people.
It is preposterous to suggest that technological applications and medical interventions do not require 'proof' - is there anyone reading this forum who would approve of CBT/GET treatments on the basis that the hypothesis that "CBT/GET helps some people with M.E/CFS" has not been falsified ? Somehow I don't think so, and that what would be expected is some very strong proof/evidence that CBT/GET actually does help people with M.E/CFS, before anyoe reading this forum would take part in CBT/GET. Hypothesis of disease association or causation may not of itself demand 'proof' but the application of technology in achieving a test of a hypothesis does require proof of adequate function - in the XMRV case, if the technology of observation is shown to be vulnerable to error, then validation of that technology is required - that is, the technology has to be 'prooved' as sound.
It should be noted, though, that when Cort "agrees" on the need to "wait for science to play out before making conclusions" it is little more than lip service meant to placate. It is the same kind of "agreement" that politicians can frequently be seen making on the need for accountability and transparency in government. It is the same kind of "agreement" that the participants of CFSAC have made for the past umpteen years on the need for better funding, etc.
Such "agreement" from Cort will always be immediately followed by a flurry of posts desperately attempting to do just what he agreed not to: foist a premature conclusion onto casual readers.
If you read between the lines of these posts, they scream out: How many times must I (Cort) spin the facts and all but centrifuge reality; how many times must I impugn Dr. Mikovits for her actions and not her science; how many times must Trine Tsouderos manufacture a notion of "scientific consensus" by citing the same 5 profoundly biased researchers; how many times must Kim McCleary and Suzanne Vernon refer to XMRV in the past tense; how many press releases must the Wellcome Trust release to prop up their non-replication studies; how much already will it take to get you people to drop the XMRV thing and get back on the CAA/CDC's 50-year plan to "solve CFS"?
You are once again misunderstanding how burden of proof works. If the WPI can consistently distinguish patient samples from control samples in a blinded experiment, then it is not their burden to explain/prove why others cannot.
In such a scenario (which has yet to be conducted), there is no need for an extraordinary explanation, much less any explanation, as to why strict adherence is required. They will have validated that their methodology is sound and meaningful. In such a scenario, it is not their burden to explain the myriad forms of incompetence and/or failures displayed in the negative papers.
This is quite a peculiar comparison to make, given the obvious double-standard involved.
In the UK (and to some degree in the US), patients are denied other treatment options and exclusively given CBT/GET despite the absence of any "proof/evidence that CBT/GET actually does help people with M.E/CFS".
Compare that to your presumed concern over ARVs, the logical "technological application and medical intervention" implied if the XMRV association pans out. In this case, patients are being actively discouraged from using them supposedly because of the lack of "proof/evidence" that they will help patients with ME/CFS.
So, to summarize, in one example (CBT/GET) the "lack of evidence" is used to support the intervention to the exclusion of others, while in the other example (ARVs/XMRV) the "lack of evidence" is used to discourage the intervention.
You can also try a Google Site Search
Separate names with a comma.