Derek Lowe on how science gets done

Organic chemist Derek Lowe writes about XMRV at his blog, In the Pipeline:

…this isn’t looking good at all. It’s pretty typical, though, of how things are out at the frontiers in this business. There are always more variables than you think, and more reasons to be wrong than you’ve counted. A theory doesn’t hold up until everyone who wants to has had a chance to take some big piñata-shattering swings at it, with weapons of their choice. So, to people outside of research: you’re not seeing evidence of bad faith, conspiracy, or stupidity here. You’re seeing exactly how science gets done. It isn’t pretty, but it gets results in the end.

If you have never read Derek’s blog, you are missing his wonderful insights into drug discovery in the pharmaceutical industry.

26 thoughts on “Derek Lowe on how science gets done”

  1. Carlitos Gonzalez

    Vincent, someone called RRM, says in the comments of that post this:

    An additional, revealing piece of information that is not in the report by Science:

    In the first round of testing for the “project” (for the Blood Working Group), when two labs found XMRV, the blood was also collected by the scientists that had reported an asociation with XMRV.

    For the second round of testing, an independent phlebotomist visited the patients to collect new blood samples, and then all three labs found no XMRV.

    Do you happen to know if that information is true?

    That is actually a relevant fact with 2 possible lectures:

    a)WPI brought contamination in the samples. (although that would not explain that one of the 3 labs was not able to find XMRV)

    b)The “independent collector” is not so independent. (But then again they would call me stupid for thinking on conspiracy)

    I appreciate your insight, if you have any on this…

  2. With medicine, we’re often at and beyond the limits of what should be considered ‘science’… but still with every researcher attempting to claim the credibility of science. For a condition like CFS, where there’s little clear-cut evidence to go upon, and lots of confident claims being made, this can lead patients to develop a scepticism towards science which is inappropriate for situations like with XMRV, where we can be confident that compelling evidence one way or the other will soon be discovered.

  3. Where I see evidence of bad faith is not in the published science itself, but in the misrepresentation of the conclusiveness of any one piece of evidence by the researchers themselves, and the subsequent echoing and amplication of those misrepresentations by the press and the Internet.

    Such, I believe, was the case with the press release issued by the Wellcome Trust, which went so staggeringly far beyond the hypothesis presented (and so far unverified by other workers) in the Hue paper. To paraphrase Mark Twain, reports of the death of the XMRV/CFS hypothesis have been greatly exaggerated. It would help if the scientists themselves were not so eager to help write the premature obituaries.

    Early science is messy and contradictory as hell and can get ugly; scientists, and even some informed laypeople, understand this, and can even see rigorous debate as a good and healthy sign. When information about the messy goings-on get filtered to the outside world via the press, however, either the mess is made to seem like mistakes, confusion and lack of progress (which we ought not to be wasting money on), or the mess is de-emphasized in favor of tidy “conclusions” that are often supported by privileging the claims of one side in a debate that may have multiple sides. (And if conclusions have been reached, we shouldn’t be wasting money on further investigation.)

    We have seen how eagerly and uncritically the Wellcome Trust press release was carbon-copied into dozens of press outlets, blogs, Tweets, and spread around the world unchallenged. Why is this a concern, in the particular area of ME/CFS research? Because reporting premature conclusions in this matter serves to reinforce an old and very damaging narrative about CFS: that the cause is impossible to find; that repeated efforts to find the cause have all been blind alleys; that ME/CFS is a ‘vague and ill-defined’ illness, with the implication that an organic cause can never be found – and then we’re back in the “hysteria” wastebasket, with the century-old ideas of Charcot and Freud to weigh us down.

    Public opinion affects support for research in very real terms; so does the official stance of the public health agencies. We have already seen several decades in which the NIH officially considered ME/CFS to be a form of depression or conversion disorder – based upon no particular hard evidence – and instructed the press to treat it accordingly. So there were few to protest when ME/CFS research was underfunded, or when funds intended for biological research were diverted elsewhere, or when grant applications for what little money was available were routinely turned down by ‘expert’ panels with no expertise in the disease. The NIH is showing encouraging signs of changing course in recent years, but we have yet to see the result in terms of hard dollars: is there any other disease as serious and disabling as ME/CFS that recieves so little NIH funding?

    http://report.nih.gov/rcdc/categories/#bpopup

    The recent article that appeared in Science magazine, though reasonably fair overall, ended on a rather sour note, which seems to me a pretty clear indication of how negative perceptions inside and outside the research community can start to lead to a chilling effect on research:

    http://www.sciencemag.org/content/331/6013/17.full

    “As the new [NIAID/Lipkin XMRV] study gets started, some wonder whether it’s worth the $1.3 million it will cost. Jonathan Stoye of the MRC National Institute for Medical Research in London concedes that the Towers study was “over-hyped.” But he says “it’s pointing people in a certain direction,” away from chasing an elusive link to XMRV. Still, he says, a larger study may be the only way to satisfy [CFS] patients.”

    The journalist’s favorite gang of straw men, we all know, is “some.” Who are these “some” people who are already suggesting that $1.3 million is too handsome a sum to spend to get good answers on XMRV, in the face of these supposedly damning recent publications on contamination? Stoye seems to be saying that the only reason to proceed with the study, in his opinion, is to shut up the patients. This implication does a disservice to the NAID’s motives in ordering and funding the study. If XMRV truly does cause or contribute to human disease, then it’s not just the problem of a group of querulous patients who wants answers – it’s everybody’s problem, and NIH doing due diligence to look for those answers is the proper way to serve the public interest. To suggest otherwise does, indeed, smack to me of bad faith.

    Thanks, as always, to Professor R. for serving as a voice of sanity amid the cacophony.

    (Apologies if this appears more than once – this Disqus system has an unfortunate habit of eating comments.)

  4. I think Coffin’s IAP assay is a red herring, because it doesn’t make sense that the results would be due to mouse contamination. For starters ,the results were originally replicated in labs which had never done any murine research. But given the sequence similarities, a human cell line source is far more likely source of contamination.
    That said, we cannot yet rule out the following hypothesis either.
    http://www.frontiersin.org/virology/10.3389/fmicb.2010.00147/abstract

  5. The info on the Blood Working Group Phase 2 that you describe are partially correct. The slides from a presentation of the BWG studies provide more detail: http://www.cfids.org/webinar/slides-121710.pdf

    Phase 2a – Samples from 4 XMRV+ individuals identified by WPI and 1 pedigreed negative control were collected and processed by WPI. Non-blinded panels were analyzed by WPI and CDC, and a blinded panel was tested by NCI. WPI and CDC got positive results, NCI results were negative.

    Phase 2b – Samples from same 5 subjects were collected by phlebotomist and sent to BSRI for processing. BSRI distributed blinded panels to WPI, CDC, NCI and GenProbe. Plasma and PBMC samples were negative by CDC, NCI, and GenProbe. WPI identified some positives, but also found the pedigreed negative sample as positive. WPI identified a procedural error that explained the false positive. There were whole blood samples in Phase 2b as well, but those results have not been presented.

    I encourage reviewing the slides because my description does not include all the relevant details. More factors besides the use of an independent phlebotomist are involved.

  6. Coffin thinks there are mouse nucleic acids contaminating common lab
    reagents, such as calf serum used for cell culture or phosphate
    buffers. So the lab does not have to work with mice for such
    contamination to occur. Human cell lines as a source is a possibility,
    but this does not explain why Lombardi et al found antibodies in human
    sera that react with XMRV. The paper you cite is interesting – I had
    planned to write about it.

  7. Richard Jefferys

    According to those slides, it looks like the antibody test used in the Science paper produced a positive from a pedigreed negative sample. It also appears to have produced different results from day 0 and day 2 samples from a putatively infected person from the Science paper.

  8. Wouldn’t culturing your samples change the assertion regarding antibodies? When you have contaminated your sample and then culture it (culturing, although not described in the Lombardi et al paper, is what happend to get to those 67% PCR results), I feel it’s pretty logical that antibodies can form? Or isn’t this the case?

    Also, Groom et al (one of the negative studies) found antibodies in 14% of healthy control and in 0% of CFS patients, while NCI (Ruscetti) did detect antibodies in the Blood Working Group sample that from a healthy patient and was pedigreed as negative by WPI/NCI/CDC for PCR/serology/culture beforehand. Based on those findings, it seems the “serology proves we’re right” argument doesn’t seem to hold up well

  9. Pingback: Tweets that mention Derek Lowe on how science gets done -- Topsy.com

  10. If I understand your comment – you are suggesting that antibodies are
    forming in cell culture? That does not happen – antibodies would only
    arise in humans, except in special cases. As for the Groom study – I
    think we need more studies to reach a consensus. There is enough
    uncertainty so far to warrant this.

  11. I agree that the Wellcome press release was problematic. From my experience here with such releases, it is likely that the Wellcome Trust conferred with Towers over the wording (his work was funded by the Trust). He likely approved the final version, whose conclusion is quite different from that stated in the paper – which is much more vague. It’s unfortunate that this release was echoed by much of the world’s press without looking at the original papers.

    I just had a conversation with Dave Tuller (author of several NY Times articles about XMRV) which I will post as a TWiV episode in the coming weeks. We touched on the issue of press releases and XMRV, as well as the mixing of science and journalism.

    I do appreciate your concern about research being influenced by such negative comments. However, be assured that a good part of basic virology research is investigator-initiated. If someone wants to work on XMRV, they get some data and write a grant proposal. It is reviewed by virologists (in this case, retrovirologists) and if it is meritorious it will be funded. The decision should not be based on what is written in press releases, but rather what is in the primary scientific literature. While some virologists are publicly negative about XMRV as a human pathogen, I believe they would judge such grant proposals on their scientific merit.

  12. Yes, you did understand my comment. It didn’t seem unreasonable to me, but thank you very much for the clarification.

  13. “While some virologists are publicly negative about XMRV as a human pathogen, I believe they would judge such grant proposals on their scientific merit.”

    Have you seen who makes up the Center for Scientific Review: Special Emphasis Panel for CFS?

  14. Gob, ‘would like to know who does make up the panel. I’ve heard in recent past, James Jones was responsible for blocking WPI’s proposals. JJ is welcome to his PNI perspective on his own time and health; we can’t have that crap keep happening. (I’ve read an account of his lecturing the Tahoe victims in the ’80s, assuring them they all had something in their lives they’d wish they’d done differently. Yah – tell that to those “abused” kids.)

    To Dr. Vince: ME/CFS are supposed to have more say this year in the grant process. In the past, NIH stonewalled much CFS research, aside from Syndey Grossberg, who unlike DeFritas stayed out of the limelight. Per Oslers Web, even an epidemiologist who’d been tracking HIV was turned down. I, and I believe Gob, don’t believe this was all due to science.

  15. I wish I could share your belief in the unbiased nature of grant funding. However, the WPI has submitted 10 grant proposals to the government, none of which have been funded. I am confident all had scientific merit, and that the decision to not fund was political.

  16. The last two panels looked a bit better, but the virologists who were on them this time could not vote on virology papers because of their involvement in MRVs.

    So at the last meeting Hanson & Ruscetti could not vote.
    Fillingim & Friedberg are psychologists.
    Jones works for the CDC and has long held ME/CFS to be psychosomatic.
    And then there is Broderick, Fletcher, Klimas, Biaggioni, & Szomolanyi-Tsuda. None are virologists.

    Great situation!

    Here are two links to the last meetings of the SEP CFS:

    http://www.csr.nih.gov/Roster_proto/meeting_roster.asp?stdate=6/22/2010&enddate=6/23/2010&grcode=80&srg=CFSM&SRGDISPLAY=CFS-M&agsqnum=226610

    http://www.csr.nih.gov/Roster_proto/meeting_roster.asp?stdate=11/2/2010&enddate=11/3/2010&grcode=80&srg=CFSM&SRGDIS(PLAY=CFS-M&agsqnum=232163

  17. The first question is whether the sample was a negative. With improved testing it may be shown not to be.

  18. Unless you have read all 10 WPI grant proposals, you should not make
    conclusions about their scientific merit. Producing a ‘fundable’ grant
    proposal is no simple feat these days – I have seen dozens of superb
    grant applications fall by the roadside. Grant funding is not unbiased
    – but I do believe that in the long term the science gets done. This
    conclusion is based partly on the fact that have watched and
    participated in scientific grant application and assessment process
    for nearly 30 years.

  19. hi profvrr….cant we read the actual grant applications? is there somewhere we can read them….I would be interested in your take on them, it would be interesting becuase if the concensus was that they were top notch applications and they were turned down then it would throw up some interesting questions of WHY they have been turned down, could be an interesting exercise? is it possible?

  20. As far as I know, NIH grant applications are treated as confidential
    documents and therefore are not available to the public. You can find
    out which grants have been funded at http://RePORT.nih.gov/ (use the
    RePORT tool) but you will only find the abstracts there.

  21. Richard Jefferys

    For this suggestion to be the explanation, it would require that one supposedly infected person changed from positive to negative in the space of two days while the pedigreed negative person went from negative to positive in the space of two days. This is not plausible.

  22. Justin Reilly, esq.

    Prof. R.,

    I totally defer to your opinion on the grant process in general as you have a great deal of experience. However, we informed ME patients, researchers and clinicians are the experts on whether the NIH and CDC grant processes for ME studies specifically have always been extremely biased. The answer to that question is, absolutely beyond cavil, Yes. This is a matter of well documented fact. Please defer to us in this matter (unless and until you study the specific question of ME funding- which i encourage) as we are the experts.

    For the first several years, NIH made no grants on ME despite many reportedly excellent proposals. Jay Levy, Seymour Grufferman, Klimas, Komoroff etc were turned down. Meanwhile for the first 15 years CDC misappropriated most of the ME budget and lied to Congress year after year about it; they also insisted when asked by the congressional funding committee if they wanted more funding, they said no, that would be inappropriate. NIH and CDC have generally funded studies attempting to show psychogenesis of ME even though this has long been disproved. Science will reveal the truth in the end, but if that is after say 50 years, this disease will have spread unchecked in that time period causing a monumental amount of suffering and death. That is unacceptable.

  23. Prof. R.,

    Here is required reading for anyone before they declare that the there isn’t significant bias and bad faith how NIH, CDC and the british government has dealt with ME including funding. It is an excellent, brief op-ed piece in the NY Times by the authority on this matter- Hillary Johnson who wrote the exhaustive 700 page history on the federal govt’s war on ME science and patients, Osler’s Web.

    http://www.nytimes.com/2009/10/21/opinion/21johnson.html?_r=1&scp=1&sq=hillary%20johnson&st=cse

Comments are closed.

Scroll to Top