DecodeME blog: X marks the spot where ME/CFS biology can be discovered

Simon M

Senior Member (Voting Rights)
Shortened version– read the full blog here.
This is from an earlier version - please check against the published blog

Scientists, people with ME/CFS, and their charities came together to create DecodeME, the world's biggest ME/CFS study – and its results are striking. 18,000 people with ME/CFS gave their DNA, enabling DecodeME to reveal eight genetic signals for the illness. These signals indicate that immune and neurological processes play a significant role in ME/CFS.

Why a DNA study?

We know little about the biology of ME/CFS, which makes it very hard to develop effective treatments. Usually, we can’t tell if the biological differences studies find between those with ME and those without are a cause of the disease, or an effect. DNA is different because diseases don’t change our DNA.

DNA carries the instructions for how to make and run a human body. It’s effectively a very long string of four different chemicals, named A, T, C and G for short. Some of these letters make up our roughly 20,000 genes, and each gene carries the recipe for making a particular kind of protein. Proteins are our bodies’ ‘doing’ molecules, such as digestive enzymes, antibodies and components of muscle fibres.

DNA-science.png


As most proteins do something very specific, if we can pinpoint which genes are different in people with ME/CFS, we also pinpoint the protein involved – and that points to the biology.

How DecodeME looked for DNA differences

DecodeME was a genome-wide association study (GWAS). These studies don’t look for DNA differences in each of the three billion DNA letters that humans have, because that’s very expensive. Instead, they look only in those places where people often have different letters from each other (each different letter at a particular location is a ‘variant’). DecodeME looked at nearly nine million of the DNA letters, but that is still only three in every thousand letters.

A diagram of dna and snp  AI-generated content may be incorrect.Caption: Illustration of a DNA variant where one person has DNA with the letter A, while another has the letter G at the same position.


DecodeME’s findings

DecodeME looked at the DNA of 15,579 people with an ME/CFS diagnosis who met the study’s tight research criteria and compared it with the DNA of nearly 260,000 controls without ME/CFS…

The analysis revealed eight genetic ‘signals’ for ME/CFS from the nine million variants studied. And these signals will help to reveal the biology of the disease.

You can see these signals as the peaks (‘skyscrapers’) that stick up above the dotted line in the ‘Manhattan plot’ below. They show where there are groups of variants that are significantly different in people with ME/CFS compared to healthy people. These peaks, or signals, are each flagged with the name of the gene that could explain them.

1754504409320.png

Image: [Manhattan plot for the main DecodeME GWAS showing 6 of the eight genetic signals. Two that were just below the threshold were above it in two other GWAS using DecodeME participants – see blog]


The genetic signals are like crosses on a treasure map, pointing to hidden biology.

GWAS are an affordable way to scan the human genome but are harder to interpret because they only look directly at a few DNA letters in every million. This means that although there are eight genetic signals of closely-grouped variants significantly associated with ME/CFS, the DNA defined by the signal will contain many unseen variants. These unseen variants are more likely to be a cause of ME/CFS because there are far more of them.

So each genetic signal is like an ‘X’ on a treasure map indicating roughly where the researchers should dig for treasure. The next step is to find that treasure – the specific genes that are causing ME/CFS.

DecodeME identifies top genes
DecodeME has started searching for the treasure, and has identified likely genes using several methods. These methods looked at genes within the genetic signals – the peaks on the Manhattan plot.

… see the blog for more details, but the most likely genes mainly relate to problems tackling an infection, plus neurological processes including pain.

…So, DecodeME showed that immunological and neurological processes are likely to play a part in ME/CFS. They also found a link to chronic pain,

DecodeME also looked at genetic signals associated with depression and anxiety that were near the genetic signals found for ME/CFS

So, DecodeME showed that immunological and neurological processes are likely to play a part in ME/CFS. They also found a link to chronic pain, but not to anxiety or depression.

Researchers could flood in to transform the field
The real story here is that DecodeME has generated a host of new biological clues.

Sonya Chowdhury, chair of DecodeME’s management group and CEO of Action for ME, said, ‘These results are groundbreaking. With DecodeME, we have gone from knowing next to nothing about the causes of ME/CFS, to giving researchers clear targets.’

She added, ‘We hope this attracts researchers, drug developers, and fair funding to ME/CFS – and speeds up the discovery of treatments.’


Read the full blog
 

Attachments

  • Manhattan plot master 22 7.png
    Manhattan plot master 22 7.png
    450 KB · Views: 3
  • DNA-science.png
    DNA-science.png
    4.1 KB · Views: 3
  • DecodeMe-SNP.png
    DecodeMe-SNP.png
    12.3 KB · Views: 5
  • Manhattan plot master 22 7.png
    Manhattan plot master 22 7.png
    450 KB · Views: 13
Last edited:
Thank you for a very informative summary!
Another gene on the plot, SUDS3,produces a protein that dampens down the inflammatory response of the brain’s main immune cells, called microglia. So people with ME/CFS could have a weaker microglial inflammatory response in the brain.
Wouldn’t that be ironic if we actually have less brain inflammation?
DecodeME used another analysis method called MAGMA to link ME/CFS genetic signals to the nervous system. MAGMA first identified a set of 13 genes associated with ME/CFS signals. Then, using existing data on gene activity, it found that this set of genes was more active than expected in 13 types of tissue in the human body (out of a possible 54), and all 13 tissues were brain regions.
That seems like a strong indicator that the brain is central in ME/CFS?
 
I’ve done a quick pass at audio of the two blog posts and the abstract from the paper to help people access the results. They’re available at the below link. I’ve assumed this is okay with the authors, if not let me know and I’ll remove them.

https://u.pcloud.link/publink/show?code=kZqX2W5Z4yR8YkHhWUzzghs07AaMamsl6mW7

I’ll look at cleaning up any problems and get to the rest of the paper later as that’s a bigger job. If anyone has the straight text or even a latex file of the paper please message me, pdftotext conversion always seems to introduce extra text formatting problems to account for.
 
Enormously grateful to the DecodeME team and all the researchers, advocates and patients involved. This is such fundamentally important work that can now be built on.

The entire process-as a patient-was very thoughtful in design—with regular updates and information, even including the email ‘heads up’ to pace for the impending results. Much appreciated.

I really look forward to the science that will be driven forward by this study’s results!
 
DecodeME used another analysis method called MAGMA to link ME/CFS genetic signals to the nervous system. MAGMA first identified a set of 13 genes associated with ME/CFS signals. Then, using existing data on gene activity, it found that this set of genes was more active than expected in 13 types of tissue in the human body (out of a possible 54), and all 13 tissues were brain regions.
What does "brain regions" mean here? Meaning specific parts of the brain? Or does that cover the entire nervous system?
 
Thank you, @Simon M. Clear and informative as always. :thumbup:

DecodeME was a genome-wide association study (GWAS). These studies don’t look for DNA differences in each of the three billion DNA letters that humans have, because that’s very expensive.
Any idea how expensive, and how much value there is in a full detailed gene map study?

Or do we have enough now to start focusing down?
 
Any idea how expensive, and how much value there is in a full detailed gene map study?

Or do we have enough now to start focusing down?
I think the ambition is to sequence 10,000 people, and the cost would be I think at least £6 million (sequencing costs have fallen a lot since DecodeME was funded five years ago).

The proposed new study already has a name – SequenceME

It works in a different way to GWAS. Certainly, it would make identifying the relevant genes much easier, because you have a full sequence, not a few variants and lots of gaps.

But its main value is fundamentally different. A common difference is that GWAS look at 10 not to be harmful – otherwise they wouldn't be common. Mostly they involve small difference in gene expression brackets the amount of protein that is produced from the gene, and sometimes the timing or location of it)

By contrast, whole genome sequencing, WGS, can spot faulty genes – where the gene code is for a faulty protein. These are the problems we see in classic mal genetic diseases like muscular dystrophy and sickle cell anaemia

But there will be people with illness who have something similar – but only very small numbers and you have no chance of picking them up in a study like this. WGS tend to text these faulting jeans, which are often stronger clues.

DecodeME has strengthened the case for SequenceME, because we can now show there's something to be found.
 
Just wanted to echo all the thanks and congratulations to everyone involved in the project.

But also say what a great blog post this is from @Simon M and thank him for it. I’ve been able to share it with lots of people I know who aren’t deep into ME/CFS, science or even that aware of DecodeME because it’s so accessible and covers a lot of the background as well as the findings in a concise way.

We really do have some great science communicators in the community.
 
Wouldn’t that be ironic if we actually have less brain inflammation?
According to AI:
in the findings is SUDS3,produces a protein that dampens down the inflammatory response of the brain’s main immune cells, called microglia. So people with ME/CFS could have a weaker microglial inflammatory response in the brain. Does this mean less inflammation or more in the brain


Great question—and it's a bit nuanced. If the SUDS3 gene produces a protein that dampens the inflammatory response of microglia (the brain’s resident immune cells), then people with ME/CFS who have this genetic variant may have a weaker microglial inflammatory response.

So does that mean​

Possibly yes—but not necessarily in a good way. Here's why:

Microglia and Brain Inflammation​

  • Microglia are crucial for responding to infections, clearing debris, and maintaining brain health.
  • When overactivated, they can cause neuroinflammation, which is linked to symptoms like brain fog, fatigue, and pain.
  • But if their response is too weak, the brain may not properly deal with infections or cellular stress, leading to chronic dysfunction without classic inflammation.

What SUDS3 Might Imply​

  • If SUDS3 dampens microglial activity, it could mean less overt inflammation.
  • However, this might also result in ineffective immune surveillance in the brain, allowing low-grade infections or cellular stress to persist.
  • That could contribute to the chronic neurological symptoms seen in ME/CFS, even without high levels of inflammatory markers.
In short: less inflammation doesn’t always mean better. In ME/CFS, it might mean the brain’s immune system isn’t doing enough, rather than doing too much.
 
Thanks so much, @Simon M, without your blog I wouldn't really understand any of it.

It's great to hear that the results are what we were all hoping for: an A road with destination signs, after spending 40 years lost in an industrial estate.

Congratulations and sincere thanks to everyone who's worked on this project. The way it's been executed from Day 1 has been exemplary.
 
Great blog @Simon M . Thank you.

Perhaps someone can help my confusion. I see references to these 8 aberrant genes "causing" ME/CFS. Is that literal? I mean, do these genes create what we call ME/CFS? Do they make us more susceptable? Does everyone have them, but they need to be"switched on" by something in order to cause ME/CFS? What are the implications to the cause concept? For example, does this disprove the persistence theory, or does it instead make it more likely?

Sorry if that is unclear. :(

Clarity is not my strong point.
 
Great blog @Simon M . Thank you.

Perhaps someone can help my confusion. I see references to these 8 aberrant genes "causing" ME/CFS. Is that literal? I mean, do these genes create what we call ME/CFS? Do they make us more susceptable? Does everyone have them, but they need to be"switched on" by something in order to cause ME/CFS? What are the implications to the cause concept? For example, does this disprove the persistence theory, or does it instead make it more likely?

Sorry if that is unclear. :(

Clarity is not my strong point.
Not an expert, but I can try my best to explain.

Some of the genes are just slightly less or more common in me/cfs, and thus tell us something about what precesses are involved in developing me/cfs or staying ill. But none of them just straight up cause me/cfs, or guarantee someone to either get it or not.
 
Back
Top Bottom