Preprint Initial findings from the DecodeME genome-wide association study of myalgic encephalomyelitis/chronic fatigue syndrome, 2025, DecodeMe Collaboration

These comments are from the media coverage thread but I’m posting here as they are relevant to the study design and interpretation of data:

Maat said:
Being covered on Sky news now. Only the Psychologist on the panel was asked to comment. She stated that what people are pointing out is that many of the participants in the study had 'self reported diagnosis'. Am I correct in recalling that only those with a medically confirmed diagnosis were invited to provide samples, or am I confused? I do remember that I had to wait to see if I would be called to provide a sample.
Click to expand...

Simon M said:
There were two parts to this. The study was of people who reported they had a diagnosis from a health professional. The study screening questionnaire also asked people about their symptoms. Only people who met the criteria for either IOM or Canadian consensus definitions were invited to provide DNA samples. I'm not sure, but I think about 85% of people who signed up saying they had a diagnosis from a health professional also met the criteria.

DecodeME included a question on postexertional malaise that went beyond asking about simple exertion intolerance stressing the need for an extending period of symptom flare. Though, as this is a media thread, I don't want to start an extended debate about that here.

But I want to clarify this isn't self diagnosis, it's not even simply self report of a medical diagnosis. It won't be perfect, but I think it will be pretty good.
Click to expand...
Given that the study has found genetic associations, does the fact that the study used self-reported symptoms not strengthen the result? I can see that it was an unavoidable weakness in the study design, which could have been used as an explanation if it had found no associations but, to my limited understanding, finding these associations despite the unavoidable weaknesses in the design, suggests that these signals could be stronger and more numerous if there was a way to screen patients more thoroughly (eg examined and tested by physicians as part of the study).

Please correct me if I’m misunderstanding.
 
Last edited:
Sasha said:
So they'd do all the analyses, which would take six months, and then they'd submit it, so that publication could take quite a bit longer?
Click to expand...
The end of August marks the end of our funded period, and we will not get further funding from the MRC and NIHR. This means that almost all of our team then move on to other projects elsewhere, while Chris and a very small number of people supported by relatively small amounts of philanthropic funding continue the work.

The remaining analyses may, or may not, take up the estimated 6 months, there is no certainty, and then peer reviewed publication takes as long as the peer review process takes.
 
Robert 1973 said:
These comments are from the media coverage thread but I’m posting here as they are relevant to the study design and interpretation of data:




Given that the study has found genetic associations, does the fact that the study used self-reported symptoms not strengthen the result? I can see that it was an unavoidable weakness in the study design, which could have been used as an explanation if it had found no associations but, to my limited understanding, finding these associations despite the unavoidable weaknesses in the design, suggests that these signals could be stronger and more numerous if there was a way to screen patients more thoroughly (eg examined and tested physicians as part of the study).

Please correct me if I’m misunderstanding.
Click to expand...
I agree, it can be argued that the signals could be stronger if you have say something like a clinical cohort, but I do think it's equally fair to argue that there is also the alternative explanation: It is due to confounding factors (for example comorbidities).

I think the larger difference is here that people wanted to partake in a single ME/CFS study. I think that naturally might introduce different confounders than say Lifelines where as I understand it you are just part of cohort.

I think what these people are trying to argue: "Well these people just have bit of anxiety, maybe some FND or some mix of other issues and GWAS there also find some things whilst it's still BPS."
 
ME/CFS Science Blog said:
The failed replications are disappointing but I suspect that next to broad case definitions for ME/CFS in the other databases, the sample size might also have been too small.
Click to expand...
Yes, the replication wasn't a complete bust. It does look like bigger samples could give significant results.
the preprint said:
Four of the eight loci (RABGAP1L, FBXL4, OLFM4,CA10) were associated at p < 0.05 with cases ascertained using post-exertional malaise and fatigue in the UK Biobank and the Netherlands biobank Lifelines.
Click to expand...

(I find the Results Replication section a bit hard to follow. I can't quite work out what has been done and what, if anything, is still planned.)
e.g.
For our UKB replication attempt, we plan two GWAS using cases and controls from the UKB, and created GWAS-2, a new DecodeME GWAS (described above), as an appropriate comparison for them.
Click to expand...
 
Jonathan Edwards said:
I think we are on fairly safe ground there - with caveats. At least one gene is pretty specific for a subgroup of T cells - not even any old T cells. T cells will be upstream in this because T cells don't feel ill, brains do. So it looks like it is going to have something to do with immune cells making brains feel ill. That could still be via screwing up metabolism in other places.
Click to expand...
Which gene is this? If you’re referring to BNT2A2 that’s not true, it was first identified in mammary cells after all. Also in antigen presenting cells and other non-immune cells.
 
Wow, so many of the candidate genes are immune - I really hope this generates some interest among immunologists.

There are also numerous mentions of lupus in the candidate gene document:
RC3H1 "Lupus-like autoimmunity"
TNFSF4 "Region upstream of TNFSF4 is associated with systemic lupus erythematosus risk"
TRIM38 "A minority of lupus and Sjogren’s syndrome patients have autoantibodies reactive to TRIM38"
 
Fizzlou said:
What is happening with a US DecodeME cohort? Was that Lipkin and is his funding back on?
Click to expand...
I am not sure but I don't believe that we know at the moment due to the chaos in the US, even though in theory Columbia's funding has been restored.

InitialConditions said:
Apologies if I've missed this, but do we know which journal this has been submitted to?
Click to expand...
As there is still work to be done it has not yet been submitted to any journal.
 
EDIT: these probably reflect artefacts that were excluded in the main analysis?

I'm looking at the DecodeME summary data, namely the file: gwas_1.regenie.gz taken from here: https://osf.io/rgqs3/files/osfstorage

When I arrange by lowest p-value however, I get 8320 rows with a p-value lower than 5 * 10^-8. Some go all the way up to 2.44^-62.

1754571256851.png

The Manhattan plot looks like this. Is there something that I'm missing here or is there an error in the summary data file?



1754571155308.png
 
One of the things I'm wondering is what next. Given the genes of interest that have been found what are the next steps to understand potential mechanisms; what would help test the hypotheses; and which would be good research groups in the right areas?

Is there some low hanging fruit that could be easily tested?
 
Adrian said:
One of the things I'm wondering is what next. Given the genes of interest that have been found what are the next steps to understand potential mechanisms
Click to expand...

I guess ideally it's SequenceME?

I wonder if by trying focusing all efforts on the important genes we'd risk only looking under the lamppost. They look relevant to the story, but there could be rarer variants that fill it out better or say more about specific pathways.

There's likely a lot more that can be done with the data we already have, though.
 
You must log in or register to reply here.
Back
Top Bottom