Lipkin's Monster ME/CFS Study: Microbes, Immunity & Big Data
The Microbe Discovery Project outlines an ambitious new study by top researchers that has collected patient samples, but needs desperately funds to complete the work.
Discuss the article on the Forums.

The Power and Pitfalls of Omics: George Davey Smith’s storming talk at ME/CFS conference

Discussion in 'Phoenix Rising Articles' started by Simon, Feb 15, 2016.

  1. msf

    msf Senior Member

    Messages:
    3,197
    Likes:
    4,484
    Well, the definition of linkage disequilibrium seems to include cases (theoretical?) of the latter, at least according to Wikipedia:

    In spite of its name, linkage disequilibrium may exist between alleles at different loci without any genetic linkage between them and independently of whether or not allele frequencies are in equilibrium (not changing with time).[
     
  2. msf

    msf Senior Member

    Messages:
    3,197
    Likes:
    4,484
    That was just an example. If you think about it, linkage disequilibrium must exist between non genetically related genes, since some genes will improve the chances of other genes´ survival.

    It seems to me that at the very least mendellian randomization must control for ethnic origin, since if two mutations in two different genes arose in one population, and then that population mixed partially with another population, there would be linkage disequilibrium between those genes in the combined population.
     
    Invisible Woman likes this.
  3. MeSci

    MeSci ME/CFS since 1995; activity level 6?

    Messages:
    7,969
    Likes:
    12,814
    Cornwall, UK
    There's some stuff about linkage here.

    For example:
     
    Simon and Invisible Woman like this.
  4. anciendaze

    anciendaze Senior Member

    Messages:
    1,809
    Likes:
    4,666
    Has anyone told George Davey Smith about the political nightmare associated with this disease, and the problem of getting any uniform cohort of 10,000? Is he aware that ignoring dynamic response of individuals to exercise as "obviously psychological" or statistically insignificant will cause the disease to disappear, even if you are studying the right people? A large number of patients are perfectly well equipped to spend the remaining years of their lives lying down. By some common medical thinking this means they are healthy, except for this peculiar choice of how to spend their time.
     
  5. Simon

    Simon

    Messages:
    1,921
    Likes:
    14,549
    Monmouth, UK
    Which earns you a Geek award:
    [​IMG]
    Not sure: unless the mixing was very recent, Mendelian assortment would break the association.

    Don't forget, only a relatively small proportion of genes will be in LD with the gene of interest in a MR any case; unless those few genes also affect the outcome (say, coronary heart disease for Vitamin C), it wouldn't matter. Since very, very few genes will be in LD with the "MR" gene, it's not too likely to be an issue.

    But maybe this calls for an MR expert, or some papers on the subject.
     
  6. msf

    msf Senior Member

    Messages:
    3,197
    Likes:
    4,484
    Yes, I´m a geek!

    It doesn´t just depend on how recently the two populations came together, but how much mixing goes on within the combined population, which is determined by mate selection and so on. I´m pretty sure that if you didn´t control for race in a genetic study of the population of the USA, there would still be a linkage disequilibrium between the alleles that predispose someone to sickle-cell anemia, and those that increase the amount of melatonin in the skin. Of course, you may view 400 years ago as fairly recent, but most populations are a more recent mixture than that!

    I agree that it is unlikely that the allele for high availability of Vitamin C will be in linkage disequilibrium with an allele that decreases the risk of coronary heart disease, but I think the fact that it is a possibility is one of the reasons that UptoDate states that Mendellian randomization can not replace the double-blind placebo-controlled trial.
     
  7. Sasha

    Sasha Fine, thank you

    Messages:
    12,789
    Likes:
    34,216
    UK
    I think we should just all pause for a minute and gloat about having got someone of GDS's stature and expertise involved in our disease.

    Ron Davis has been saying for a while that ME/CFS is the last, big uncharted disease and is exciting territory for any scientist. If titans like GDS are entering the field now, he's being proven correct.
     
    searcher, *GG*, Valentijn and 5 others like this.
  8. MeSci

    MeSci ME/CFS since 1995; activity level 6?

    Messages:
    7,969
    Likes:
    12,814
    Cornwall, UK
    I'll try, but there are many years of suffering and head-wall-banging that need compensating. Here is the nearest emoticon I can find for gloating:
    :smug:
    Maybe we can add:
    :):cool::D:love::wide-eyed::jaw-drop::woot: (if that can be translated into a sensible word rather than 'Woot') :thumbsup::angel::nerd::balloons::star::trophy::bouquet::music::wine::bulb::fire::moneybag:
     
    Snow Leopard, Valentijn, Bob and 2 others like this.
  9. Sasha

    Sasha Fine, thank you

    Messages:
    12,789
    Likes:
    34,216
    UK
    Outstanding use of smileys, @MeSci!

    I'm very fond of Woot! :woot: but wish it was called something else. Reminds me of Fozzie Bear from The Muppets.
     
    MeSci likes this.
  10. Maggie21

    Maggie21

    Messages:
    25
    Likes:
    170
    Sounds interesting. Already looking forward to reading part 2.

    In fact already wondering how we can volunteer to take part in any resulting research . . . any one know?
     
    Simon likes this.
  11. ukxmrv

    ukxmrv Senior Member

    Messages:
    4,335
    Likes:
    4,407
    London
    What could we expect to see from this research i.e what would be the outcome?
     
  12. Esther12

    Esther12 Senior Member

    Messages:
    8,449
    Likes:
    28,523
    Thank you Simon. Interesting stuff.

    Simon and Davey Smith, sitting in a tree...

    Is that right? If some genes lead to someone being better at absorbing some (or just one) nutrients, couldn't that have all sorts of positive affects on how they develop when growing up and then go on to live their lives? If it means that they are more likely to be healthy and intelligent, that could affect all manner of adult behaviors. Or it could not. I have no idea, so am just asking. I guess that if it did have those affects, that's all the more reason to supplement nutrient X?

    When a diagnosis of 'CFS' seems so much less useful than something like 'MS' I worry that The Grand Challenge might not be as useful as some hope. It would be interesting to hear any thought GMS had on how to deal with the uncertainty of 'CFS' for studies like this.

    Inevitable snark:

    Maybe he shouldn't have put his name to that letter complaining about the mistreatment poor old Wessely suffered from critics in the Independent?
     
    Valentijn and worldbackwards like this.
  13. Simon

    Simon

    Messages:
    1,921
    Likes:
    14,549
    Monmouth, UK
    Ha!

    Seems somewhat unlikely given these are rather modest effects. The effect of gene difference of say, the vitamin C transporter gene, are quite modest leading to small average group differences in vitamin C level between 'high' and 'low' groups. I can't think of any mechanism where such a minor effect will substantially affect smoking or, say, socioeconomic status. That's another reason why you need big, big samples for these approaches: you are looking at the impact of modest changes.

    It's not impossible that genes affect later behavious, of course. One neat example of Mendelian randomization uses a gene that affects the body's ability to break down alcohol to predict actual alcohol consumption and study the effects of alcohol on health (such studies were part of the evidence for recent advice that even low level drinking has a small but significant effect). But again the effects are relatively modest, even where there is a very clear reason for a link between genes and behaviour.

    Finally, I'm pretty sure that when you check the data, the MR gene of interest is well-distributed across,say, social class ie it isn't a case of people with genetically higher levels of Vit C having a higher socioeconomic status.

    Snark noted. Even so, I remain thrilled that someone like GDS wants to join the fray - and I'm sure he'll learn about the politics soon enough.
     
    Last edited: Feb 18, 2016
    Esther12 and Valentijn like this.
  14. Esther12

    Esther12 Senior Member

    Messages:
    8,449
    Likes:
    28,523
    Thanks Simon.

    Yeah, that stuff would be checked for (presumably), and any association would (presumably) raise a red flag.

    There is still a danger that the benefits of access to some additional nutrient would only be of benefit at a particular point in someone's development. Or even that it could be of benefit at one point, and harm another. I realise my concerns are starting to get a bit thin there, but I do like how a well conducted RCT lets us not worry about bizarre unknowables that could possibly affect the results of something like this. Admittedly, that is largely because I'm pretty lazy, and prefer fool-proof techniques that help me avoid needing to think about things too much: hopefully GDS has got his thinking cap on.

    Also, I've never thought about this approach before, and CFS history has given me an instinctive fear of anyone claiming to be an expert who knows anything unless I can really see that their claims are justified.
     
    Last edited: Feb 18, 2016
    merylg, MeSci and Simon like this.
  15. Simon

    Simon

    Messages:
    1,921
    Likes:
    14,549
    Monmouth, UK
    You're way ahead of the game, @Esther12

    Yup, that is one of the weaknesses of MR, not knowing developmental factors like that. But it is good evidence that, say, overall having higher Vit C across your life is a good thing.

    There's no doubt the RCT is the very best way to study these things (just look at PACE :)) but MR is enourmously useful because it's cheaper, faster and sometimes allows you to do things that would be unethical with a RCT.

    Take that vit D/MS study. The hypothesis based on epidemiological data was that vit D in childhood reduces the risk of MS. You''d have to set up a randomised study, follow people from childhood til they were, say, 45 - over thirty years - and measure MS. And it would have to be a huge study as MS is relatively rare. Also, people might not be too keen to have their kids taking a placebo when the evidence was that vit D was a good thing, which would probably sink the study alone - regardless of the thirty-year timescale and vast cost. The study was actually done quickly, cheaply and ethically, by reusing data genomic data from other studies.

    Alcohol use is another area where a RCT might be problematic (who'd take the risk of the no-alcohol group, or max alcohol...) and an MR study gives good data.

    Also, I'm summarising a very complex talk including a lot of other stuff than MR so this isn't even a comprehensive take of his talk. But there is an extensive literature on MR, including debating its weaknesses (pleiotropy is a significant one) but in many cases these weaknesses don't apply. It's like anything, needs to be done thoughfully. And do enjoy the literature.
     
    Last edited: Feb 18, 2016
    MeSci, voner, Sasha and 1 other person like this.
  16. Simon

    Simon

    Messages:
    1,921
    Likes:
    14,549
    Monmouth, UK
    In case you missed it, part 2 is ready and waiting
    The power and pitfalls of omics part 2: epigenomics, transcriptomics and ME/CFS

    Also, Davey Smith was interviewd today (March 1st), available to listen online or download
    We could definitely use that in mecfs research.
    BBC Radio 4 - The Life Scientific, George Davey-Smith

    added: somewhere in that interview he describes Mendelian randomisation (though not by name) as fiendishly complex and hard to grasp. So if, like me, you found it less-than-obvious, you are not alone.
     
    Last edited: Mar 1, 2016
  17. MeSci

    MeSci ME/CFS since 1995; activity level 6?

    Messages:
    7,969
    Likes:
    12,814
    Cornwall, UK
    I heard it, and was pretty-much in agreement with him until the end, when he showed his enthusiasm for mass medication, namely statins (but I suspect that he feels the same about some other things, e.g. by the look of his teeth - public water fluoridation?). I was rather surprised that he spoke as though statins had no or almost no adverse effects.

    It would be interesting to quiz him on the desirability/possibility of identifying those who are more or less likely to benefit from drugs of mass medication (sounds a bit like weapons of mass destruction, and maybe similarly undesirable?). Scientists are trying to identify who will and won't benefit from rituximab, for example.

    GDS is very much in the public eye at the moment, isn't he? Is he on chat shows yet?
     
    ukxmrv likes this.
  18. *GG*

    *GG* Senior Member

    Messages:
    6,161
    Likes:
    4,767
    Concord, NH
    I recall reading an article in Life Extension, and think the guy said statins was part of his protocol for health. I was surprised to read that, all I heard was it was bad for you, reduces CoQ10 and causes muscle pain for some. I think it might lower some marker in your blood, which is supposed to be good for you.

    GG
     
  19. Cheshire

    Cheshire Senior Member

    Messages:
    1,121
    Likes:
    8,999
    Bob, Simon, Scarecrow and 1 other person like this.
  20. Sasha

    Sasha Fine, thank you

    Messages:
    12,789
    Likes:
    34,216
    UK

See more popular forum discussions.

Share This Page