Preprint Dissecting the genetic complexity of myalgic encephalomyelitis/chronic fatigue syndrome via deep learning-powered genome analysis, 2025, Zhang+

SNT Gatchaman said:
FWIW I ran the top 20 genes through Gemma3:27b on my laptop. The prompt was "I'm going to give you a list of 20 genes. Find interactions to explain mechanisms that produce the disease ME/CFS."
Click to expand...
Thanks. I think you might get better results if you don't mention ME/CFS and just ask it if there are any patterns in these genes that were found to be abnormal. Otherwise it will try to connect it to popular memes in ME/CFS research such as inflammation, mitochondrial dysfunction etc.

I tried it with the 115 genes that had a p-value below 0.001 and prompted shit GPT a couple of times (in different conversations with slightly different wording) to see if it it came up with the same patterns, which was the case.

Here's a typical response:
One striking theme is the prevalence of genes involved in intracellular signaling and synaptic function, including NLGN1, NLGN2, DLG2, DLGAP1–4, SYNGAP1, GRM1, CAMK2A, and HOMER2. These genes are closely associated with postsynaptic density and play essential roles in neurodevelopment and synaptic plasticity. Many of them have been implicated in neurodevelopmental disorders such as autism spectrum disorder and intellectual disability, particularly CHD8 and SYNGAP1.

Another major cluster involves oncogenic and cell cycle regulatory pathways, notably genes like DNMT3A, DNMT3B, PIK3CA, NRAS, NOTCH1, RET, CDC6, CDC23, BUB3, HDAC1, E2F6, and SMARCD3. These are commonly mutated or dysregulated in cancers, including hematologic malignancies and solid tumors. Their abnormalities suggest either a proliferative or epigenetic disruption in affected tissues.

There is also a group tied to metabolic regulation and mitochondrial function, including NAMPT, AK2, AK3, PANK1, PANK2, PANK3, GALT, COASY, CA2, and PPCDC. These genes regulate energy metabolism, cofactor biosynthesis, and redox states, hinting at altered metabolic homeostasis, which can be relevant both in cancer and neurodegenerative conditions.

Genes such as INS, LEP, and ADCY10 point toward involvement in endocrine signaling and homeostasis, possibly implicating insulin signaling, glucose metabolism, and neuroendocrine integration.

A fourth group involves immune function and inflammatory response, with entries like IL12A, HLA-C, STAM2, and NFATC3. These suggest dysregulation in immune signaling pathways, which may overlap with inflammation-driven oncogenesis or autoimmune phenomena.

Finally, there’s a surprising overrepresentation of proteasome components and protein degradation machinery, such as PSMB3, PSMB4, PSMB5, PSMD7, PSMC3, and PSMC5. This supports a pattern of disrupted protein turnover or stress responses, which again can tie into either neurodegenerative diseases or cancer biology.
Click to expand...
 
Sasha said:
If a small, deep-learning study spits out lots of statistically significant genetic associations, do we take them seriously or not? If we have to wait for better information, is this information not reliable? And if it isn't, what's the point of doing such studies?
Click to expand...
No one study anywhere in science stands alone. Even if it's not as robust as a study with 20,000 people, it still adds to the weight of the evidence, and seeing replications between DecodeME and this study would strengthen the findings. Even a giant study will likely have some meaningless findings come up due to chance. Seeing the same findings in different populations using different methodologies decreases the likelihood of that being the case for those genes, and helps prioritize research directions.
 
An important point is that the HEAL2 algorithm leverages the STING database to incorporate protein-protein interactions in its attention mechanism. What this means for interpretation is that multiple genes that are all part of the same network are likely to score higher cumulatively. Therefore, having more genes in the same pathway should not necessarily be taken as evidence of the pathway’s relevance above other pathways.

But it is good information if multiple pathways converge towards a similar point, especially if it recapitulates other findings in the field. To that end, I’m particularly interested in what signaling pathways get upregulated in response to viral infection and, in healthy people, get methylated/deacetylated after infection.

I think the story will also have to involve glutaminergic [edit: signaling (neuronal or immune)], calcium signaling, cAMP, and metabolic regulation (leptin) at some level. That makes it a very broad biological search space, but still better than where we were before.

Overall, due to the focus on rare loss-of-function variants, and the graph network in HEAL2, I think this methodology is much less likely to output a fruit salad of significance like early GWAS studies. The main concern would be bias towards genes in large well-characterized protein interaction networks.

However, I think this is more likely to result in many genes being ignored by the model, rather than a high false positive rate. Given the long list of genes that already came up, I think that’s a preferable problem to have.

Like I’ve already said, I don’t think these findings are definitive, but they do offer useful information to the extent that they can be cross-referenced with what has already been found in the field and what will hopefully come out of DecodeME.
 
Last edited:
If we will only have confidence in these results if they're confirmed by different forms of analysis that we can be confident in, I don't see what information these results are adding. Isn't this the very definition of confirmation bias?
 
Simon M said:
But until we know that’s the case, I don’t think we should take it on trust. I gather the AUC for the replication cohort wasn’t very impressive here, which doesn’t inspire confidence.
Click to expand...
It was nearly the same as their AUC on their training cohort, which is impressive in itself. I would lose confidence if it was a high AUC in training and near 0.5 in test.

While weaker, the training AUC is about what I’d expect from a rare variant analysis on a smaller cohort.
 
Sasha said:
If we will only have confidence in these results if they're confirmed by different forms of analysis that we can be confident in, I don't see what information these results are adding. Isn't this the very definition of confirmation bias?
Click to expand...
Not necessarily, I think the point of a rare variant analysis is to reaffirm that loss of function in those pathways is in fact critical for disease pathology.

It would be a similar logic as doing a knockout study in a mouse. If you have experimental results that suggest a gene is involved in disease pathology, the next step would be knocking it out in a mouse and seeing whether that induces the disease under certain conditions.

That’s not to say this is equivalent to a mouse study—obviously that should come later once we have basis for disease models. But I still think it’s useful corroboration. If we had several important pathways from experiments but none of them were coming up in genomic studies, that would be a more worrying problem in my opinion, as it suggest that those pathways are extraneous to the disease.
 
jnmaciuch said:
and what will hopefully come out of DecodeME.
Click to expand...
I think it’ll be great when these findings are published because they will finally give us a fairly solid reference point. At the moment, it’s not that hard to find study results to support most theories. Thesse reference points will make much of the literature more interpretable.
 
Last edited:
jnmaciuch said:
An important point is that the HEAL2 algorithm leverages the STING database to incorporate protein-protein interactions in its attention mechanism. What this means for interpretation is that multiple genes that are all part of the same network are likely to score higher cumulatively. Therefore, having more genes in the same pathway should not necessarily be taken as evidence of the pathway’s relevance above other pathways.
Click to expand...
Interesting point. Not speaking from expertise or experience but I would think these networks are sufficiently complex so that it still means quite a lot if multiple genes from a pathway are highlighted.

I suppose that having an abnormal result for one gene highlights the pathways it is involved as being more likely. But there are likely countless and numerous ways it could be connected to other genes and pathways. So if various genes in for example synaptic function light up, I think this reliably shows that this pathway is more likely to be relevant. There is likely some reinforcement (i.e. gene A in pathway 1 highlights gene B in the pathway 1 and vice versa) but that is probably the only way to get significant results out of such a small sample size.
 
Simon M said:
I gather the AUC for the replication cohort wasn’t very impressive here, which doesn’t inspire confidence.
Click to expand...
On the other hand: the AUC is measured against current diagnostic practices of ME/CFS which may not be very precise anyway in terms of pathology. Suppose only a small subgroup has pathology involving synaptic function, then the maximum AUC score would be quite low.

So perhaps what matters most in this context is that it seems to capture a (modest) signal, in that it could separate patients from controls using both 5-fold cross-validation and an independent cohort.
 
Last edited:
ME/CFS Skeptic said:
Interesting point. Not speaking from expertise or experience but I would think these networks are sufficiently complex so that it still means quite a lot if multiple genes from a pathway are highlighted.
Click to expand...
They are quite complex, however they're highly biased by known interactions in the literature. If nobody thought it was interesting to check if protein A binds to protein B, it’s not going to end up in the database even if it’s very relevant to the disease. And the algorithm has a cut off for number of interactions, so it’s going to be heavily biased by what has been extensively studied already. It’s the same across all biology—something that has already been well characterized continues to get more attention simply because it has already been well characterized.

So I agree that these pathways are relevant, and your point about finding signal in a small sample size is exactly the author’s justification for using it. I just caution against using the number of genes in a pathway as a proxy for gauging importance in disease. It may very well be that something like synaptic remodeling is related to the pathological mechanism of ME/CFS, but by several degrees of separation.

And as Jonathan already pointed out, many many of these genes are doing double or triple duty, often in similar systems, so while we can say that e.g. proteins involved in glutamate signaling come up repeatedly, the connection to synapses is one of inference rather than fact.
 
Just on the discussion of the SequenceME project size:
Simon M said:
But I’d be surprised if they got funding for as many of 17,000.
Click to expand...
https://www.actionforme.org.uk/sequenceme-first-of-a-kind-genetic-study/ Dec 2024

The ambition and what is possible given funding seems to be different. There's an intent to analyse 17,000 samples, but it seems likely that funding will limit that.

The partners are working together to secure funding for a study which will analyse the entire genetic code of up to 17,000 people with Myalgic Encephalomyelitis (ME) in a bid to uncover the genetic causes of the illness.
Click to expand...
Over 17,000 participants who donated saliva samples to DecodeME have consented to further analysis and he SequenceME partners will seek to analyse them all.
Click to expand...
This quarter, the study partners concluded a successful pilot phase by completing any-length sequencing of ten individual samples from the DecodeME library, demonstrating the high accuracy and scalability of the study method. The next phase, involving sequencing of 10,000 participants, requires £7 million in funding.
Click to expand...

Here's the thread on SequenceME:
SequenceME genetic study - from Oxford Nanopore Technologies, the University of Edinburgh and Action for ME
It would be good to get an update on the funding situation.
 
Sasha said:
If we will only have confidence in these results if they're confirmed by different forms of analysis that we can be confident in, I don't see what information these results are adding. Isn't this the very definition of confirmation bias?
Click to expand...

The confidence comes from the combination. Think of it like two people listening to a piece of music, one through the wall of a concert hall and the other on a really crummy radio with interference. The first person says 'I am. pretty sure it's Beethoven from some of the harmonies but I really can't tell which one. The other person says 'There is definitely singing as well as orchestra so it is either Das Leid von Der Erde or Matthew Passion or Beethoven's 9th.

So it's Beethoven's 9th.
 
Jonathan Edwards said:
The confidence comes from the combination. Think of it like two people listening to a piece of music, one through the wall of a concert hall and the other on a really crummy radio with interference. The first person says 'I am. pretty sure it's Beethoven from some of the harmonies but I really can't tell which one. The other person says 'There is definitely singing as well as orchestra so it is either Das Leid von Der Erde or Matthew Passion or Beethoven's 9th.

So it's Beethoven's 9th.
Click to expand...
That analogy only holds up if we can be certain that we're getting a true signal from both sources that's merely degraded by noise. I think we can take a standard GWAS or WGS analysis as providing a true signal plus noise but do we know that about this machine-learning technique? Could it b simply rubbish plus noise?
 
Sasha said:
That analogy only holds up if we can be certain that we're getting a true signal from both sources that's merely degraded by noise. I think we can take a standard GWAS or WGS analysis as providing a true signal plus noise but do we know that about this machine-learning technique? Could it b simply rubbish plus noise?
Click to expand...
My layman’s understanding is that the machine learning model doesn’t add info. The algorithm it applies might be rubbish, but it doesn’t change the underlying data.
 
You must log in or register to reply here.
Back
Top Bottom