EU Horizon funding - Prof. Simon Carding, €7.5 million

[Jonathan Edwards]

Why worryingly? Bigger samples sizes and more omics data that is not hypothesis-driven seem quite welcome.

There is no harm in getting more data but when I have reviewed grants on more omics I have looked for new approaches or techniques that are presented as pivotal to success. We are not given any specific here. Big samples run into the probem of statistically significant findings due to low level systematic bias. I wonder whether we are going to see anything much more beyond the Edinburgh omics result. That produced a few signals but maybe due to selection bias and very non-specific. The proteins I thought might be specifically interesting have not really been confirmed elsewhere i think and Chris was sceptical of their relevance anyway.

Without the causal power of genetic studies I think we are relatively unlikely to see major clues coming out of more omics when we already know that there are no sore thumbs sticking out in smaller samples. Maybe there is more here than just omics but the blurb does not say that much.

 

[butter.]

The consortium might not be able to do everything on our wish list but it will enable large omics studies which will likely be useful for the entire field. So don't quite get why that would be worrying.


There might be new leads if we test ME/CFS with large enough sample sizes.

These Horizon Europe calls are also more suited to large collaborations that are multidisciplinary take everything on board (there are lots of requirements and checklists to be met). So applications focused on a particular topic like genetics are more difficult to get funded with this grant.

It’s worrying for several reasons. You can’t obtain meaningful data readouts if the samples are not properly pre-stratified. Patients may be at completely different stages of PEM, have blood drawn at different times of day, different stages of disease, are on diets and so on, all of which makes it far less likely to detect clear signals.

Why? Because if there were strong, obvious differences that did not depend on proper stratification, we likely would have identified them already. The size of the sample helps, but only marginally, I would say. It would be much more sensible to focus on a better biobanking framework and then run the omics.

EDIT: One more thing, to run eg metabolomics on such a 'mixed sample' is more or less useless, I am afraid. I hope I am wrong.

 

[ME/CFS Science Blog]

You can’t obtain meaningful data readouts if the samples are not properly pre-stratified. Patients may be at completely different stages of PEM, have blood drawn at different times of day, different stages of disease, are on diets and so on, all of which makes it far less likely to detect clear signals.

The biobanks will likely record the most important stratifiers and harmonize their methods. Not sure if it will take everything you mention into account but it will likely be better and more robust than most of the studies we had thus far. And with big enough sample sizes abnormaliities limited to subgroups can be sufficient to detect a difference. For example if only 10% of patients is in PEM during sampling and a pathway is only disrupted then, this might be become a group difference if you have enough data.

Big samples run into the probem of statistically significant findings due to low level systematic bias.

Think that's mainly just an issue of interpreting the results correctly. The estimates will be closer to the true result compared to the smaller studies that we had before and with less variation. That will make it more difficult to claim a significant finding when there isn't one.

I see no downside to collecting more ME/CFS data. The EU is paying, not the ME/CFS community.

I wonder whether we are going to see anything much more beyond the Edinburgh omics result. That produced a few signals but maybe due to selection bias and very non-specific.

My impression is that the omics studies we had thus far were so small that it was basically all noise a bit like genetics studies before DecodeME. Think it's worth setting up something bigger. I agree that null results are the most likely outcome but those might be useful as well.

The Edinburgh study used ME/CFS cases from the UK Biobank which had problematic selection criteria. But if something comes up in both this HE consortium and the Edinburgh study that would be quite interesting.

 

[Casterofspells]

It’s worrying for several reasons. You can’t obtain meaningful data readouts if the samples are not properly pre-stratified. Patients may be at completely different stages of PEM, have blood drawn at different times of day, different stages of disease, are on diets and so on, all of which makes it far less likely to detect clear signals. None of this is currently controlled for in biobanking.

Why? Because if there were strong, obvious differences that did not depend on proper stratification, we likely would have identified them already. The size of the sample helps, but only marginally, I would say.

EDIT: One more thing, to run eg metabolomics on such a 'mixed sample' is more ore less useless, I am afraid. I hope I am wrong.

The fact we have multiple omics type studies will still help though. If signals reappear we have a better handle on whether they are significant or caused by some bias. Together with the German national decade against IACCs funding a long term biobank it’s maybe not the most effective, cumulative way of working one would wish for but it should give us comparable data points at least.

 

[butter.]

The biobanks will likely record the most important stratifiers and harmonize their methods. Not sure if it will take everything you mention into account but it will likely be better and more robust than most of the studies we had thus far. And with big enough sample sizes abnormaliities limited to subgroups can be sufficient to detect a difference. For example if only 10% of patients is in PEM during sampling and a pathway is only disrupted then, this might be become a group difference if you have enough data.


Think that's mainly just an issue of interpreting the results correctly. The estimates will be closer to the true result compared to the smaller studies that we had before and with less variation. That will make it more difficult to claim a significant finding when there isn't one.

I see no downside to collecting more ME/CFS data. The EU is paying, not the ME/CFS community.


My impression is that the omics studies we had thus far were so small that it was basically all noise a bit like genetics studies before DecodeME. Think it's worth setting up something bigger. I agree that null results are the most likely outcome but those might be useful as well.

The Edinburgh study used ME/CFS cases from the UK Biobank which had problematic selection criteria. But if something comes up in both this HE consortium and the Edinburgh study that would be quite interesting.

I am not saying the study has no worth (apart from the metabolomics maybe), but I'd say it's clear that improving biobanking first would have been sensible.

PS: I have read the original biobanking papers and criteria a few years ago, I was not impressed. If you can check them out.

 

[Casterofspells]

Any more info on what this 7.5 million Discover-ME study is planning on doing?

That delay makes me wonder what's going on with SequenceME funding we don't know about yet...

What do we reckon of the chances of EU Horizon funding for Sequence?

This project sounds interesting, I hope it covers more than just the usual microbiome stuff.

Sequence ME funding from this would be amazing. That might really move the needle. But I don’t know if a UK based project is still permissible for those funds at all?

 

[Jonathan Edwards]

Sequence ME funding from this would be amazing.

I suspect that SequenceME just wouldn't fit into this multi-group template.

 

[butter.]

The fact we have multiple omics type studies will still help though. If signals reappear we have a better handle on whether they are significant or caused by some bias. Together with the German national decade against IACCs funding a long term biobank it’s maybe not the most effective, cumulative way of working one would wish for but it should give us comparable data points at least.

Sure. Securing this funding is a huge success, but please consider the implications: if this 7.5 million multiomics study fails to yield real results, obtaining future funding for similar omics investigations will be exceptionally difficult. There is a real risk that the underlying sampling and pre-stratification are the issues, rather than a lack of biological signal there, though. While I cannot quantify this risk, it is a real concern. So much depends on better biobanking, I believe. This will be really difficult to pull off.

 

[Kitty]

These Horizon Europe calls are also more suited to large collaborations that are multidisciplinary take everything on board (there are lots of requirements and checklists to be met). So applications focused on a particular topic like genetics are more difficult to get funded with this grant.

You'd think genetics could be included as one of the elements in a multidisciplinary, multi-centre approach though? When trying to unpick a disease like ME/CFS it's obviously going to be one of the key avenues.

 

[Casterofspells]

You'd think genetics could be included as one of the elements in a multidisciplinary, multi-centre approach though? When trying to unpick a disease like ME/CFS it's obviously going to be one of the key avenues.

The news just got out a few minutes ago from Germany that they will have a genome sequencing project as well powered by Germany’s biggest pre existing Helmholtz databank.

Nationale Dekade gegen Postinfektiöse Erkrankungen: Erste Maßnahmen festgelegt - BMFTR

Es geht voran bei der Forschung gegen postinfektiöse Erkrankungen wie ME/CFS oder Long-Covid.

www.bmftr.bund.de

 

[forestglip]

The news just got out a few minutes ago from Germany that they will have a genome sequencing project as well powered by Germany’s biggest pre existing Helmholtz databank.

Nationale Dekade gegen Postinfektiöse Erkrankungen: Erste Maßnahmen festgelegt - BMFTR

Es geht voran bei der Forschung gegen postinfektiöse Erkrankungen wie ME/CFS oder Long-Covid.

www.bmftr.bund.de

Auto-translated section mentioning genetics, with bolding added:

The representatives of the steering committee agreed on three measures to be implemented promptly at the start of the decade:

• A comprehensive database on post-infectious diseases will be established for research purposes. To this end, the Federal Ministry of Finance, Research and Innovation (BMFTR) is funding the sequencing of the genomes of healthy and infected individuals to identify potential biomarkers, risk factors, and protective or preventive factors. As a first step, the sequencing will be carried out using existing cohorts from the NAKO Health Study and the Network of University Medicine (NUM).
• To scientifically evaluate the success of potential therapeutic approaches, the conduct of clinical trials will be strengthened. Corresponding funding measures will be published shortly.
• Initially, three working groups will provide the steering committee with topic-specific support: One working group will focus on translational research, which aims to translate clinical research findings into practical applications. Another working group will address research infrastructures, data, and biospecimens. A third group will focus on research relevant to healthcare provision. It is possible to establish further working groups or subgroups.

To enable the scientific evaluation of potential therapeutic approaches, the conduct of clinical trials will be strengthened. The steering committee and working groups of the National Decade against Post-Infectious Diseases will meet regularly in the future to develop the long-term strategy of the National Decade against Post-Infectious Diseases. Further measures are to be initiated later this year.