Development and validation of blood-based diagnostic biomarkers for [ME/CFS] using EpiSwitch®… 2025, Hunter et al. (Oxford Biodynamics)

[Ash]

I am well aware that many people think that way but the more people with ME/CFS and advocacy organisations for ME/CFS call for and fund projects focused on this sort of diagnostic test the longer the disease will remain in limbo with researchers churning out useless analyses of what might otherwise have been valuable data.

At least Robert seems to understand that !

Look at inflammatory rheumatic disease - we went from total ignorance to having highly effective treatments for all of them without ever bothering with a diagnostic
Click to expand...

I was agreeing with you.
Are you disagreeing with me or….?

 

[Jonathan Edwards]

Hmm. We have been homing in on particular signals like interferons (1 and 2) and DecodeME has pulled up CA10, OLFM4 and BTN2A2 etc. - without any presuppositions about cell types in either case, just disease dynamics and DNA data.

This study picks out IL-2, CD4, maybe HLA-DR, IL-10 in PBMCs, which are all about things like IL-2, CD4, HLA-DR and IL-10. It's a bit like saying we looked at white cells and found a major focus on white cell gene products.

IL-2 could be relevant, as an indicator of T cell activity. But we have already found studies that show that physical activity shifts T cell behaviour.

The comparison of disease and drug 'networks' smells to me of AI fairyland stuff - which we have seen before and not known what on earth to make of. Give me an old-fashioned GWAS any day.

 

[Jonathan Edwards]

was agreeing with you.
Are you disagreeing with me or….?

Ah! Right. I read your post backwards!! It must be my rhinovirus, or just old age.

 

[Utsikt]

The issue with this test is very simple in practical terms:
There are loads of reasons for why someone would be flagged by this test that has nothing to do with their ME/CFS.
There are also loads of reasons for why someone with ME/CFS would not be flagged by this test.

This is mainly caused by two factors:
The lack of a closely matched control group and no grounding in biological processes known to be causally involved in ME/CFS.

Therefore, this test can’t ever validated a diagnosis.

Using this test to validate a diagnosis is just as nonsensical as using CPET to validated the disability for disability benefits, or using a tachycardia upon standing as a diagnostic feature for OI in general or the (very flawed concept of) POTS.

Edited for clarity.

 

[Ash]

Ah! Right. I read your post backwards!! It must be my rhinovirus, or just old age.

Or a lack of predictive capacity on my part ‘cause…

 

[poetinsf]

There is no point in developing a diagnostic test that correlated 100% with a syndrome. If it does, you know you already had a 100% method of recognising the syndrome so the test is unnecessary.

It could still be useful if we are moving from the diagnosis of exclusion, which often takes years, to a positivity test that takes days. This particular test, however, smells like a marketing job. For all we know, they might have had exactly the same result if they mixed ME/CFS cases with flu cases.

The pathways section looks to me just like a random bucketful of the usual suspects.

This was exactly my first thought, a marketing job.

 

[ChronicallyOverIt]

View attachment 28721

To be fair lots of low cap biotech are ripping today, and the market in general though not 40%…. Though the amount of press for this is for sure a pump.


They were trying to fundraise beginning of this year they seemed to be in a bad position. I would not be surprised if this is a last ditch effort to garner some capitol.


Proposed Fundraising - 16:35:00 14 Jan 2025 - OBD News article | London Stock Exchange
https://www.londonstockexchange.com/news-article/OBD/proposed-fundraising/16852409

 

[InitialConditions]

This study doesn't warrant all this media attention. Wonder if the SMC is to blame for pushing this?

I really doubt it. It's probably due to PR from Oxford Biodynamics.

 

[SNT Gatchaman]

Then it wouldn’t be a replication. Replication is to do the same thing in a different lab by a different team and seeing if it gives the same result.

Is that the case? I thought it meant replication of results, but allowed for variation in methodology, including up to completely different methodology. The idea is to make the findings more robust.

 

[Simon M]

Is that the case? I thought it meant replication of results, but allowed for variation in methodology, including up to completely different methodology. The idea is to make the findings more robust.

The open science movement, which was the 1st to push replication hard as a solution for life science problems, has defined it broadly to include both direct replication of methods and trying to replicate the overall findings more generally. The first approach includes taking the original method and improvingo upon obvious weaknesses, such as a small sample size. .

 

[Utsikt]

I think that was the note that I finished with, I understand the requirement.

I was just wondering since medicine often doesn’t meet such requirements, why this particular category specifically was problematic.

I’ve read a fair amount about POTs and I still don’t understand what it’s supposed to be exactly.

I’ve also read a fair few people dissecting POTs, and dismissing it, I think that where I’d stand.

But, I struggle to retain any of this information. I’m not in disagreement with your initial statement, I really couldn’t remember the info on this category.

I don’t really care too much about what’s become standard. I only care about what’s logically sound. And POTS is not. I just used it as an example of similarly bad practice.

Is that the case? I thought it meant replication of results, but allowed for variation in methodology, including up to completely different methodology. The idea is to make the findings more robust.

I was under the impression that replication was about trying the same thing to see if it gave the same results. But that might be «replication of the study» and not «replication of the results». So I can see that there might be room for a different understanding.

I think that if you were to fix all of the weaknesses of this trial, it would be so substantially different that it can’t reasonably be called a replication of the original study.

And I’m not sure what such a study would be looking to replicate. Would they try to use the same biomarkers in new cohorts? Or is it the analysis to identify the biomarkers? Both? Something else?

 

[DMissa]

In terms of specific functional implications I’m not sure what we can draw from this study because it is mixed PBMC

 

[Jonathan Edwards]

In terms of specific functional implications I’m not sure what we can draw from this study because it is mixed PBMC

Right, and since the molecules picked out are cell type specific it makes interpretation particularly fraught.

 

[MelbME]

It's annoying that they took the "it's a diagnostic test" approach to the data when simply comparing ME/CFS to healthy people. You don't need a test to know they are not a healthy population, you need to know why they aren't a healthy population, what is it that is causing this and what markers distinguish it from other diseases/conditions?

That aside, it's nice to see some of the immune hits they got. It's encouraging that they did training, test and validation.

 

[DMissa]

Right, and since the molecules picked out are cell type specific it makes interpretation particularly fraught.

I should have mentioned that this was my reasoning!

 

[DMissa]

That aside, it's nice to see some of the immune hits they got. It's encouraging that they did training, test and validation.

One thing is that if we take immune cells and then the pathway hits are broadly immune related that isn’t really saying much, since it’s what would be expected by chance.

 

[jnmaciuch]

Looking at the methodology, this seems like exactly the same information that a Hi-C assay would be able to give you, but even less so since it’s only looking at chromosome “loops” and not general structural proximity. I had a chance to attend a lecture on Hi-C a few months ago by someone who has put out a lot of studies since the technology was developed, and it was one of the first times I’ve heard a scientist admit that the last decade of work with their method of choice has by and large failed to deliver on the promised insights. I can’t imagine that proprietary Hi-C Lite would do much better.

Interpretation-wise, a chromosome loop just tells you that there are probably more CTCF linkages in the vicinity of certain genes, which is itself probably downstream of any number of transcription factors. It might indicate that something is being transcriptionally regulated in that general area but even in that regard, a run-of-the-mill ATAC-seq would probably give you more specific and useful information.

From prior chromatin studies I’ve worked on, I can confirm that a massive amount of chromatin remodeling happens in myeloids just as a result of normal aging. It’s a pretty safe bet that age and comorbidities are going to be driving a huge amount of the variance here—if those weren’t controlled for, I wouldn’t even spend time speculating on any genes that may have popped up here.

 

[Hutan]

Posts about POTS, OI and POT have been moved to POTS Definition, Diagnosis and Symptoms