What research do you want to see? (study ideas)

forestglip said:
Why not throw in an XXX as well? In theory, the risk of ME/CFS might be higher still if there is a triple dose. That's what seems to be the case in one of the SLE studies.
Click to expand...

Maybe, but one can argue that the situation for XXX raises a different question about repression - of a third X as well - that isn't strictly relevant to the difference between one (male) and two (female) in the way that the others are.
 
I imagine the simplest way to get this dosage data would be to compare the proportions of these with the various X-chromosome make-up spelled out above in the DecodeME cohort versus the general population. Although I don't know if DecodeME collected such data.
 
Simon M said:
Although I don't know if DecodeME collected such data.
Click to expand...

My worry would be that people with XO or XXY might be less likely to offer their DNA for studies, or conceivably more likely. There could be a major skew and the numbers would be quite small. On the other hand if there are any data they might just about be interpretable. Of course it is possible that people with unusual karyotypes were excluded?
 
Jonathan Edwards said:
Of course it is possible that people with unusual karyotypes were excluded?
Click to expand...

The data analysis plan for DecodeME says:
Sex of participants are inferred during the automated genotype calling process in AxAS based on X and Y linked variants. Samples failing this inference are flagged as “unknown” sex. This can reflect underlying sex-chromosome aneuploidy and mosaicism. Such conditions can be identified after conducting a Copy Number Variant (CNV) analysis using the AxAS. The probeset intensities across whole sex chromosomes of “unknown” sex samples are visualised and compared to those of male or female references (7). Samples with an “unknown” sex that remain unresolved or presenting sex-chromosome aneuploidy will be flagged and removed. Additionally, samples showing a discrepancy between the self- reported sex in questionnaire at recruitment and the genetically inferred sex are also removed as indicative of potential sample mix-ups.
Click to expand...

I would assume that whatever quality control tool they used would log the details of each sample that was removed. So maybe it would say "sample 123 was removed for having the XXY genotype". Thus maybe it would be as simple as counting how many XXY's were removed and dividing by the total case count to get prevalence.

Though I do agree there's a possibility of recruitment bias one way or the other with individuals that have these genotypes.
 
The best chance of a biomarker are the problematic antibodies as we saw in what Dara did to the antibodies of responders as measured by Tyler.

Problem is, we don’t know which antibody is the problem
 
I would like to propose as a research target the function of efferocytosis which this hypothesis suggests that is a central causal mechanism for Post Exertional Malaise (PEM).

The table below shows identified targets relevant to efferocytosis :



Screenshot 2026-06-18 at 14.28.51.webp



Using machine learning methods, efferocytosis was identified in 2017 and was circulated to a number of ME/CFS researchers.


I also had the opportunity to present my work to the MIT Lab and Michal Tal (efferocytosis was presented as a hypothesis) :

https://s4me.info/threads/machine-learning-assisted-research-on-me-cfs.5015/page-10#post-665842

More on efferocytosis on the following posts :

https://s4me.info/threads/charting-...026-hoel-fluge-mella.44410/page-4#post-678885

https://s4me.info/threads/machine-learning-assisted-research-on-me-cfs.5015/page-11#post-671938
 
I think we are now at the point where researchers should conduct small, meticulous clinical trials of the following drugs, which would function as therapeutic experiments investigating various immune mechanisms that might be contributing to a brain immune signalling loop.

Campath - suggested by JE on several occasions. Broad T cell depletion would test whether the T cell hypotheses are on the money.

Anti GDT cell monoclonal - would give some concrete answers about whether gamma delta t cells are involved, which I was told would be incredibly hard to find direct evidence for in another way.

JAK STAT inhibitor - Baricitinib or perhaps Rinvoq - obviously being looked at in long covid but a small well selected study in MECFS could provide important clarity.


There are a few more 'theory-testing' drugs that we might conceivably want to trial, but imo these three have the strongest rationale currently. I was interested in anti IFN monoclonals but iirc @jnmaciuch thought they might not be effective even if her theory is correct. It would also be very nice if there was a therapeutic drug or two that could probe the 'pure' brain side of things that has been emerging lately.

If anyone else has any opinions on drugs which have some rationale for a therapeutic experiment do chime in!
 
Last edited:
Do we have a thread on what investigations could be done on the genetic data we already have (or will have in the future)?

Various people have ideas they’ve tried themselves, I’m sure others have ideas of things they’d want to explore more of if they could. Our limitation is rarely ideas but often ability. So if we could get a clear list together maybe we could get researchers interested in doing some of them?

I’m thinking particularly of computational studies rather than clinical/pharmacological as these may have a lower barrier entry see comments by @chillier here.

We’ve seen MAGMA, FLAMES and FUMA used, some discussion of tools like SusieR, other tools like HOMER. We’ve got and can find examples of approaches used for other GWAS datasets. There’s also some more custom approaches some of us have tried like using clustering algorithms, etc. I’m also interested in if we could use AlphaGenome rather than GTEx or the human protein atlas as data sources for predicting patterns of expression and so on.

Is this thread right for this or would another one work better for this subset?
 
hotblack said:
Do we have a thread on what investigations could be done on the genetic data we already have (or will have in the future)?

Various people have ideas they’ve tried themselves, I’m sure others have ideas of things they’d want to explore more of if they could. Our limitation is rarely ideas but often ability. So if we could get a clear list together maybe we could get researchers interested in doing some of them?

I’m thinking particularly of computational studies rather than clinical/pharmacological as these may have a lower barrier entry see comments by @chillier here.

We’ve seen MAGMA, FLAMES and FUMA used, some discussion of tools like SusieR, other tools like HOMER. We’ve got and can find examples of approaches used for other GWAS datasets. There’s also some more custom approaches some of us have tried like using clustering algorithms, etc. I’m also interested in if we could use AlphaGenome rather than GTEx or the human protein atlas as data sources for predicting patterns of expression and so on.

Is this thread right for this or would another one work better for this subset?
Click to expand...
A think it would be useful with a separate thread if it’s just about ideas for ways to use the DecodeME data directly.
 
You must log in or register to reply here.
Back
Top Bottom