1. Patients launch a $1.27 million crowdfunding campaign for ME/CFS gut microbiome study.
    Check out the website, Facebook and Twitter. Join in donate and spread the word!
ME/CFS and Beating the Clock
For Jody Smith, the ticking of a clock was enough at one time to chase her back to her bed. But with the passage of time, she has been able to reclaim her living room ...
Discuss the article on the Forums.

99/101 CFS patients in SCIENCE Paper had XMRV infection

Discussion in 'XMRV Research and Replication Studies' started by DysautonomiaXMRV, Jan 27, 2010.

  1. Dr Mikovits XMRV talk: Q&A Session

    XRMV in 33 PCR Negative CFS Patients from Science paper

    19/33 Antibodies in Plasma
    30/33 Transmitable Virus in Plasma
    10/33 Protein Expression (Latent Virus)


    Which means....

    Add these to the people who were PCR positive in the same Science paper

    99 out of 101 people in SCIENCE Paper had XMRV which equates to 10 to the minus 35 chance of having CFS and not having XMRV in this study.

    NB: Editor of SCIENCE paper took out 'Virtually Impossible' to not have had CFS without having XMRV - in the study in the Science Paper and made the WPI write

    :eek: Then..... ;)
  2. Dainty

    Dainty Senior Member

    Messages:
    1,353
    Likes:
    516
    Washington State
    Just to clarify, the study was done with CFS patients who matched the Canadian criteria for CFS, correct? It seems those statistics would be expected to hold true for those who meet the Canadian criteria and might possibly be different for those who do not.

    Still, those are pretty striking odds. Thanks for posting it!
  3. Anika

    Anika Senior Member

    Messages:
    135
    Likes:
    1
    U.S.
    Virus Culture 30 plus 68 PCR is 98? where is 99th?

    At first I couldn't figure this out. Then I realized the positive PCR results (68?) had to be added, but if you add the positive culture, they only add up to 98. So, if I am counting right, there appears to be a sample that had a positive result on another test (not culture or PCR, but only another test).

    If that's right - I wonder if that is why they are saying they are now using a newer more sensitive culture test? Perhaps it would pick up all the PCR plus culture from original sample, plus the odd one out.

    Or maybe I am missing something....could well be.

    Anika
  4. Samuel

    Samuel Bedbound with NO DOCTOR

    Messages:
    221
    Likes:
    2
    Can somebody explain the probability?
  5. kurt

    kurt Senior Member

    Messages:
    1,132
    Likes:
    175
    USA.Earth
    Yea, how did that calculation work Dys? Were you talking about a probability, like a T-score maybe?

    By my figuring there was a 2% chance of not having XMRV if you had CFS in that study.
  6. cfs since 1998

    cfs since 1998 *****

    Messages:
    577
    Likes:
    4
    Dr. Mikovits isn't a statistician and got this a bit mixed up. First, the additional 31 positives were after the paper was written, and the 10^-35 was in the original paper. So the 10^35 refers to the 68/101 positives. 10^-35 is the probability that the correlation between XMRV and CFS occurred due to random chance. Basically it is "almost possible" that it is just a coincidence that most CFS patients had XMRV and most controls didn't.
  7. kurt

    kurt Senior Member

    Messages:
    1,132
    Likes:
    175
    USA.Earth
    That's what I thought, he was referring to a statistical test, which is purely a mathematical probability that groups are correlated or uncorrelated. So this does not take into account any of the science, for example, the probability that this was a false positive. That would be entirely different. And it is NOT the probability of having CFS and not having XMRV in the study, that really was 2%.
  8. dannybex

    dannybex Senior Member

    Messages:
    2,136
    Likes:
    498
    Seattle
    Huh?

    Okay, so now I'm confused.

    Whatever the 'truth' is, or the facts are...then perhaps the title of this thread should be changed if possible, to reflect that...

    ???

    thanks in advance,

    d.
  9. natasa778

    natasa778 Senior Member

    Messages:
    1,334
    Likes:
    1,063
    London UK
    All these additional tests... I am wondering if they were carried out on healthy samples. In other words can we be sure that only 3.7% of healthy individuals have the virus. How many of 96.3% of them that tested negative on PCR test would have been positive on these additional tests?
  10. fds66

    fds66 Senior Member

    Messages:
    225
    Likes:
    2
    Natasha, somewhere I saw that the controls were negative on at least some of those other tests but at the moment I don't recall where I saw that information. If I find it I will post a link - maybe someone else can remember that. It certainly wasn't in the original paper because I believe they did these tests to the patients that were negative on the original paper.
  11. bullybeef

    bullybeef Senior Member

    Messages:
    476
    Likes:
    44
    North West, England, UK
    I have had the following response to XMRV by Dr. Bell from early Nov '09. It does come across as relevant and I'll be surprised if this has never been posted before:

  12. garcia

    garcia Aristocrat Extraordinaire

    Messages:
    933
    Likes:
    96
    London, UK
    You are nearly there. Lets look at the 33 people who were PCR-negative:

    To show evidence of infection all you need is one of the above 3 measures to be positive. So how many in total of the 33 were positive? Clearly its more than 30, since 30 test positive for transmissible virus in plasma. Also it must be less than 33 since there are only 33 people. And actually the number is 31 (though it is not explicity stated anywhere), since 99-68 = 31. This means 1 person didn't have transmissible virus in plasma, but did test positive for either antibodies or protein expression (or both). In other words 31/33 of the PCR-negative group show evidence of infection.

    Think Venn diagrams.
  13. Hope123

    Hope123 Senior Member

    Messages:
    1,141
    Likes:
    531
    From the multiple versions I've heard of the WPI study, it was done with patients who matched the 1994 Fukuda criteria, some of whom also matched the Canadian criteria. Patients were hand-picked by Dr. Peterson I believe.

    Yes, I agree. Dr. M. is not a statistician and mixed up the numbers like a lot of well-educated but non-statistics oriented people do. How they calculate this is first by ASSUMING that the results are purely due to random chaos. So the chance of seeing the percentages of 68% in CFS vs. 4% in normals due just to random chaos is very, very low - the 8X 10 to the 35 number. If random chaos is very unlikely to explain the results then it is more likely the association is true and has some merit to it. Note that many studies use a p-value of 0.05 (this is 5 X 10 to the -2) as their gauge so 8 X10 to the -35 is much much better.

    The actual STRENGTH of the association is measured by the odds ratio of 54.1 -- that is people exposed to XMRV are 54 times more likely to have CFS than those who have not been exposed. 54 times is a big number.

    For p-values: http://en.wikipedia.org/wiki/P-value
    For odds ratio (OR): http://www.childrens-mercy.org/stats/definitions/or.htm

    If you understand what we've tried to explain, give yourself a pat on the back. You've done better than probably 90% of MDs.
  14. kit

    kit

    Messages:
    81
    Likes:
    0
    USA
    Thanks for explaining that Hope123. I get it now. :Retro smile:
  15. natasa778

    natasa778 Senior Member

    Messages:
    1,334
    Likes:
    1,063
    London UK
    sorry if this has been posted, but there has been lots of discussion (on this thread?) about whether they tested healthy controls for antibodies, as they did with CFS samples. just read Judy's presentation she said yes, in total about 60-70 of healthy controls - who were negative on PCR, have also tested negative for antibodies.
  16. garcia

    garcia Aristocrat Extraordinaire

    Messages:
    933
    Likes:
    96
    London, UK
    Natasa, I made this mistake too. The "antibodies" that Judy Mikovits was talking about were exogenous antibodies to viral proteins, basically added to show if there were viral proteins present.

    What we generally mean when we talk about antibodies is the human endogenous antibody response, i.e. to show whether the body has mounted an immune response to the virus. I can't find any evidence of the controls having had their antibodies tested.
  17. natasa778

    natasa778 Senior Member

    Messages:
    1,334
    Likes:
    1,063
    London UK
    Garcia, yes but isn't that the same thing that was used to double-check CFS samples? When they said that "19 of these 33 (PCR-negative CFS samples) are XMRV antibody positive" - could that mean simply that they also added exogenous XMRV antibody and this revealed that there were viral proteins present in their blood? Or did they detect endogenous antibodies in CFS sample?
  18. garcia

    garcia Aristocrat Extraordinaire

    Messages:
    933
    Likes:
    96
    London, UK
    They are talking about endogenous antibodies when they talk about the 19/33 being antibody positive.

    The antibodies for the protein are to do with the 10/33 who had Protein positive expression in Decitibine. First they added Decitibine to de-methylate, then they added exogenous antibodies to show whether the XMRV proteins were expressed.

    [​IMG]
  19. Advocate

    Advocate Senior Member

    Messages:
    506
    Likes:
    14
    U.S.A.
    Hey, Garcia, thanks for clearing that up!
  20. natasa778

    natasa778 Senior Member

    Messages:
    1,334
    Likes:
    1,063
    London UK
    thanks for clarification!

See more popular forum discussions.

Share This Page