UK: MRC and NIHR announce ME/CFS workshop, November 2019 & ME/CFS Biomedical Partnership FAQ

Chris Ponting said:
We plan to use the BOLT-LMM Bayesian mixed model association method which allows for substantial flexibility to adjust for co-factors, covariates or random effects, such as potential population structure and/or kinship among cases and/or controls.
Click to expand...

Excellent. I am pleased to hear you are not using the NUTS-GTI Dolomite Sprint version with limited turning circle.(~Sorry, a degree of levity is de rigeur here.)
 
Jonathan Edwards said:
I think it may matter because the normal genetic spread of different age groups may differ because of differential birth rates and migration. I am pretty sure that if you compared 20,000 80 year olds with 20,000 8 year olds in the UK you would find dramatic genetic differences between the populations.
Click to expand...
Yes, we will take care to investigate age-related stratification, particularly because our primary cohort (UK Biobank) is of older folk. One of the covariates used in our analysis is age too.
 
Andy said:
Many thanks for all the comments and questions. We are aiming to take note of all the feedback, and will answer all the questions we can, but that will probably take longer than we'd like simply because of all that is currently happening 'behind the scenes' - I appreciate that may well be frustrating but please bear with us for the moment.
Click to expand...
Don't put pressure on yourself Andy. All this work is very appreciated and we understand it's a lot. Most of us can't significantly contribute so any contribution matters and being able to keep at it is more important.
 
Michiel Tack said:
Most ME/CFS studies focus on patients with clinically defined confirmed ME/CFS. So patients did not simply fill in a questionnaire, they saw a doctor who probably did a lot of tests and took his time to figure out what might be causing the symptoms.
Click to expand...
[my bold]
Andy said:
2. How will the study ensure that all participants recruited to the trial really do have ME/CFS?

We take participant selection very seriously. We will be using the CureME patient questionnaire that has been in use for the UK ME/CFS Biobank for several years. This is additional to your report that you’ve been diagnosed with ME or CFS by a clinician.
Click to expand...
[my bold]

Surely they are not just relying solely on the questionnaire, but also on a medical diagnosis. Both things.
 
Barry said:
Surely they are not just relying solely on the questionnaire, but also on a medical diagnosis. Both things.
Click to expand...
Apologies if I gave the wrong impression. I did try to explain in a further post.
  • The prevalence of self-reported ME/CFS in studies is usually above 1% and I think that in some (like in the one Medfeb cites) this was true for patients who said that they were diagnosed by a clinician. It's weird that this is much higher than prevalence estimates, it could mean that the label of CFS is used for patients who would not meet diagnostic criteria for CFS.
  • Additionally, there are a couple of studies on patients where the diagnosis of CFS is suspected or made (for example by a GP) who are then examined more thoroughly at a CFS centre. These suggest a high rate (approximately 40%) of misdiagnosis.
So I don't think the problem is out of the way. My intuition would say it's still indicated to test the diagnostic procedure first to see how it compares to clinically confirmed ME/CFS.
On the other hand, I know very little about GWAS. Chris Pointing's response (thanks for answering our questions!) does reassure me a bit: if the researchers hope to find genetic signals only ever seen for ME/CFS then the risk of false positives might be less of a problem.

It's rather exciting to see such an ambitious research proposal in collaboration with the ME/CFS community. I hope it gets funded. Good luck to all those involved!
 
Michiel Tack said:
So I don't think the problem is out of the way. My intuition would say it's still indicated to test the diagnostic procedure first to see how it compares to clinically confirmed ME/CFS.
Click to expand...
My intuition is that the UK ME Biobank team's questionnaire will be a more reliable gauge of whether someone has ME than most clinical diagnoses in the UK. Many of us have been diagnosed only by GP, or by GP and ME clinic therapist.
 
Trish said:
My intuition is that the UK ME Biobank team's questionnaire will be a more reliable gauge of whether someone has ME than most clinical diagnoses in the UK. Many of us have been diagnosed only by GP, or by GP and ME clinic therapist.
Click to expand...
True but as said I do have a couple of concerns about the me biobank questionnaire (can't remember what exactly as I don't have a copy of it anymore)
 
Trish said:
My intuition is that the UK ME Biobank team's questionnaire will be a more reliable gauge of whether someone has ME than most clinical diagnoses in the UK. Many of us have been diagnosed only by GP, or by GP and ME clinic therapist.
Click to expand...
By clinically confirmed I mean by an expert clinician and according to an official case definition, as is the case in most research studies. Not a GP saying he thinks you might have CFS.
 
Michiel Tack said:
By clinically confirmed I mean by an expert clinician and according to an official case definition, as is the case in most research studies.
Click to expand...

How do you define an expert in this situation? Being a clinical expert really can only mean being able to identify an illness that is later confirmed through other means. In the absence of any other means of confirmation I am not sure how one can be more 'expert' than sticking to a set of criteria tested by questionnaire. Maybe one can but I suspect bringing in 'experts' may make the process less consistent rather than more. Patients are more likely to respond neutrally to a questionnaire than to an 'expert' who they may or may not take to.
 
Does the CureME questionnaire differentiate types of orthostatic intolerance?
NonPoTs
PoTs
Orthostatic hypotension
 
Michiel Tack said:
Someone who knows the diagnostic criteria, is qualified to test for and exclude other possible causes for the symptoms and has done this more than a couple of times in the past.
Click to expand...

I guess I would call this a competent physician rather than an ME expert.

I accept that diagnoses in GP records or given to patients have no guarantee of competence in terms of excluding other conditions. I think it would be difficult, however, to vet a questionnaire on the basis of whether or not it matched up with a result from competent exclusion. It would be hard to see how you decide that an alternative questionnaire would be better I think. Even if that was clear I think it would involve a huge lengthy research project. At some point it has to be accepted that recruiting patients cannot involve full clinical diagnostic assessment by the research team.

Again, I think that including patients who will turn out to have another diagnosis is a relatively minor problem. It will just produce a degree of dilution of the genetic signal.
 
mariovitali said:
During my visit at the CureME team in March, we tried to see whether Machine Learning was able to differentiate Healthy Controls vs PwME based on questionnaire answers. The algorithms identified a single feature that was able to differentiate HCs vs PwME with very high precision/recall (unfortunately i do not remember the exact numbers)

Perhaps we could re-visit this work so i will contact the CureME team to see if we can pursue this further.
Click to expand...

@mariovitali, I am a UK ME/CFS Biobank participant and, if I can help in any way, please shout!

Thank you @Andy @Simon M @Chris Ponting for all the info so far. It all looks excellent - if any additional samples are needed in the run up to the application, I am more than happy to provide (quick before my age means I will no longer be able to participate in the NIH study!)
 
Jonathan Edwards said:
As I understand it the healthy controls are already sorted - hundreds of thousands of them. But they are from a UK population so the patients should probably be UK.

Relatives of PWME are no good as controls for a basic GWAS screen for the reason you give. On the other hand they might be useful for other analyses. If mutations turn up in PWME it would be important to know whether those mutations tracked to illness in relatives as well. But until gene associations are found there may be little point in collecting samples from relatives.
Click to expand...

Thank you for this - I was wondering about whether it would be useful to include relatives who both/all have ME to see whether they all have the same mutations associated with disease?

Do you think there will be any attempt in this first big GWAS study (if funded!) to identify samples from patients who are related? (e.g. me and my mum both have post-mononucleosis ME, so I wondered whether it would be worth me encouraging her to participate too, and if so whether there would be a means for tagging our genetic relationship in the data).

It sounds from your comment above that the focus of this first study is to identify gene associations to the disease first, and then once these are identified, to perhaps do separate studies to look at whether these track in relatives with ME? Would it be worth including a means to tag relatives in the initial sample collection, so that even if this isn't included in the initial analysis, it could be included in future analyses without having to collect more data?

Thanks!
 
Simon M said:
Perhaps as important is the central role that patients have. Congrats (and thanks) to @Andy for his role as a co-investigator on the project.

I helped draft the Q&A, and organise the first few rounds of feedback, and I am very impressed by how willing @Chris Ponting and others have been to listen to pwME. This isn't Q&A as a way to disseminate information. It's been a process to capture patient concerns and to try to respond to them - as a way to make the project better. The plan to offer the option of paper questionnaires is just one example of this.
Click to expand...
Yes, many thanks to @Andy and @Chris Ponting, and thanks to you @Simon M for all your work and for getting Chris interested in ME research through your friendship. This would clearly not be happening without him, and goodness knows what the CMRC might be up to by now without his involvement.

Meanwhile, given the CMRC’s past connections with the SMC, it will be interesting to see if and how it covers this. I’ve not been following developments very closely recently due to additional health problems but I’m assuming that the SMC hasn’t said anything about it yet. If the GWAS study gets funding, if and how the SMC decides to cover it could have a significant effect on patient recruitment beyond social media.
 
MerryB said:
Would it be worth including a means to tag relatives in the initial sample collection, so that even if this isn't included in the initial analysis, it could be included in future analyses without having to collect more data?
Click to expand...

I am sure that the applicants have all this thought through.

Looking at which genes follow an illness within a family is a commonly used technique but it involves different sorts of analysis from a basic comparison of ill people with controls.

No doubt when the project gets up and running there will be advice on recruiting.
 
Jonathan Edwards said:
I am sure that the applicants have all this thought through.

Looking at which genes follow an illness within a family is a commonly used technique but it involves different sorts of analysis from a basic comparison of ill people with controls.

No doubt when the project gets up and running there will be advice on recruiting.
Click to expand...

Ok, thank you for your reply!
 
You must log in or register to reply here.
Back
Top Bottom