Speculations about the genetics of ME/CFS and DecodeME

I felt like writing about this.

In my opinion ME/CFS is probably a label for several different poorly understood illnesses that happen to share some characteristics and which do not yet have their own name. We might call them subtypes of ME/CFS. Then, in a ME/CFS cohort there is probably also a significant percentage of patients whose primary illness was erroneously diagnosed as ME/CFS. These could be unusual presentations of common illnesses with some similarity to ME/CFS like multiple sclerosis, or adult-onset forms of rare diseases like primary mitochondrial disorders or muscle diseases that are often difficult to diagnose.

A large study like DecodeME could have the statistical power to allow some separation of these subtypes of ME/CFS. It probably won't help the patients with the rare diseases to be correctly diagnosed. There will be a significant portion of patients that won't fit in any of the ME/CFS subtypes.

What kind of genetic variants might we find?

DecodeME is going to highlight the risk factors common to most patients with ME/CFS.

If a monogenetic form of ME/CFS was identified it could help a lot in understanding what's going wrong. My understanding is that it would require a very different kind of study, one centered on detailed genetic testing of families with clusters of ME/CFS.

ME/CFS seems to be linked to infections so one could think that it might share genetic risk factors with other infection-associated illnesses like multiple sclerosis or things like post-singles nerve pain. Maybe there are some subtle immune deficiencies or abnormal host responses to infection at play here?

ME/CFS appears to affect the brain and it might share genetic risk factors with neurological and psychiatric illnesses like parkinson's disease and schizophrenia. These could be genes that ensure healthy function of the brain.

ME/CFS also seems to have a certain distinctiveness to it, for example in postexertional malaise or how no treatments seem to work. This distinctiveness suggests a distinct pathophsyiology and this could extend to the genetic risk factors. If DecodeME highlights some genes that nobody can much sense of with no prior association to any illness, then these could be related to the more distinct aspects of the illness.

DecodeME might point us directly at genes that increase the risk or severity of PEM. These would I think be genes that are expressed in particular during the recovery phase from exercise or stressors.

I also get the impression that the triggering infection often reported by patients might be just an event that tips the balance of a system that was already vulnerable for an obscure reason. Maybe the distinction between infectious and noninfectious onset subtypes of ME/CFS merely reflect how much it takes to tip the balance towards illness. One could therefore think that the patients with non-infectious or no trigger might have more/stronger genetic risk factors (although as a whole I expect it will be a more heterogeneous group that the infectious trigger group)

 

Some more thoughts on why I think it could be genes related to recovery from exercise and stressors in general:

The fact that ME/CFS is so poorly understood could be a sign that its origin is in human physiology that is difficult to measure or maybe hasn't been studied much. These recovery processes seem like they haven't been studied much and the timeline would fit well with PEM (at least when it's delayed by 1-3 days).

Several recent studies also suggest a blunted metabolic response to exercise in ME/CFS.

If this were the key problem in ME/CFS, we might expect to see a lot of problems arising from other kinds of stressors, like infections or forced exposure to continued sensory stimulation. That seems to fit.

 

Let me make a prediction. Recently, a GWAS of long Covid found an association with FOXP4. I bet that DecodeME will find an association between FOXP4 variants and ME, at the genome-wide significance level.

Also, I think most of the genes Decode finds will be for the immune and nervous systems.

 

Possible but maybe the FOXP4 association is more with the subtypes of Long Covid that are quite distinct from ME/CFS, like hospitalized patients with severe acute illness that go on to have some degree of organ damage (in particular the lungs, which was highlighted in that paper on FOXP4).

 

ME could be due to a single gene, but it could also be a combination of many genes, varying with the individual. Just for a thought experiment, assume that ME involves an abnormal rate of production of a certain protein, which is part of a feedback loop. In one individual, 5 genes in combination change the rate of that protein. In another individual, 4 completely different genes end up having the same effect. There might be a large number of these different combinations. In a cohort of 100, you might end up with 30 groups with a correlation of a few individuals; probably not enough for statistical relevance. It will likely be worse, since of those 100, based on the not-very-good ICC, maybe 25 don't actually have ME, and many of the others are selected on criteria that represent far downstream effects of ME, meaning a genetic predisposition to ME's core dysfunction having an effect on some other mechanism.

I don't have confidence that DecodeME will pop out a clear "these two genes are responsible for ME" result.

 

We don't expect to.

 

Male/female differences in the illness might also map to different susceptibility loci on the X and Y chromosomes.

 

there doesn't appear to be a general thread on the DecodeME project?

anyway, minorish point.
on the website, why is myalgic encephalomyelitis all in lower case but Chronic Fatigue Syndrome isn't?
"We aim to find genetic causes of why people become ill with myalgic encephalomyelitis (ME) / Chronic Fatigue Syndrome (CFS) with our ground-breaking research."
https://www.decodeme.org.uk/

 

Maybe whoever invented Chronic Fatigue Syndrome trademarked it so that we have to treat it as a proper noun, like Toilet Duck.

 

Yes! I think that studying families with clusters of ME/CFS is an area that could potentially provide some good results.

In my family of origin 3 out of 6 of the children have ME and in my own family 3 out of 4 children have or had ME, so it's something I've thought about a lot.

First, you'd likely have fewer people who are misdiagnosed researching ME in family clusters as more people being sick will have had more investigation in total. If there's a symptom that isn't typical of ME that all the family members have this may be more likely to lead to an alternate diagnosis than if one individual has this symptom.

Also, assuming there is a genetic basis that is a causative factor in family clusters investigating a group consisting of family clusters could provide cleaner results as they may be in the same ME subgroup.

 

Post copied from another thread and subsequent posts moved

Decode ME saliva samples were big enough to provide DNA for a GWAS analysis and also a rare variant study (whole genome sequencing, WGS). The vast majority of participants gave consent to the rest of their DNA To be analysed by WGS. The problem is it’s very expensive, but the material is there for a very large WGS, which might be very revealing. As with all GWAS, the data also reveals family connections between participants.

 

Thanks Simon & don't feel obliged to reply to this.
Interesting --- don't suppose NIH would fund a whole genome sequence study i.e. based on the DecodeME samples!
I've noticed that using families with 2 or more members affected, and at least 1 severe, has been suggested - not clear what the evidence base for that is i.e. versus using the DecodeME approach/samples.

 

Don’t know, Sorry. I know Ian Lipkin applied for a NIH grant for a small GWAS. Not sure if this has been funded.

Also, Jarred younger, in his latest video, said he is doing a rare variant/WGS study with Liz Worthy. No mention of sample size, and I don’t recall if he mentioned who was funding work.

 

Thank you - interesting that "Ian Lipkin applied for a NIH grant for a small GWAS" - hopefully the approach means that the data could be combined with DecodeME.
Interesting re Jarred Younger/Kiz Worthy i.e. rare variant study -- yea sample size (& selection criteria) would be interesting.

 

How expensive (for a large-very large WGS)? Is anyone currently trying to get funding for it and could it happen in parallel to the GWAS? Thank you.

 

It could be £10m for a very large one. Bear in mind it was an enormous battle to win funding for a £3.5m GWAS, but I think it might happen in the medium term, particularly as sequencing prices come down.

 

Just to answer this specific part. Yes, attempts are being made to progress whole genome sequencing of the DecodeME cohort as well as attempting to secure the long term availability of data from the cohort, however nothing has been secured so far. Half of each sample donated to DecodeME was set aside and retained in the hope that we would be able to sequence the whole genome at some point in the future so, in theory, this could be done in parallel to the ongoing GWAS work but would probably need additional recruitment on the analysis side.

 

Could measures/chances to secure further funding be improved by (crowd)funding to hire people for writing grant applications, do political lobbying work and so forth? As I am quite sure the answer is yes: Are there currently existing organizational structures that could make use of such funds to implement actions swiftly?

 

In regards to DecodeME, that sort of thing would be for either the University of Edinburgh (UoE) or Action for ME, as the main charity partner, to answer; if you would like to send me an email to andy@s4me.info then I can pass it directly on.

 

That’s really interesting idea and that’s not just for a WGS. Crowd funding is a lot harder than people imagine, but the sums involved would be more doable. Though I don’t know how much fundraising appeal hiring lobbyist and people to write grant apps would have.