Re-analysis of genetic risks for Chronic Fatigue Syndrome from 23andMe data finds few remain, Bedford et al, 2021

John Mac

John Mac

Senior Member (Voting Rights)
Preprint
Abstract
It is tempting to mine the abundance of DNA data that is now available from direct-to-consumer genetic tests, but this approach has its pitfalls.

A recent study put forth a list of 50 single nucleotide polymorphisms (SNPs) that predispose to Chronic Fatigue Syndrome (CFS), a potentially major advance in understanding this still mysterious disease.

However, only the patient cohort data came from a commercial company (23andMe) while the control was a genetic database.

The extent to which 23andMe data agree with genetic reference databases is unknown. We reanalyzed the 50 purported CFS SNPs by comparing to control data from 23andMe which are available through public platform OpenSNP.

In addition, large high-quality database ALFA was used as an additional control.

The analysis lead to dramatic change with the top of the leaderboard for CFS risk reduced and reversed from an astronomical 129,000 times to 0.8.

Errors were found both within 23andMe data and the original study-reported Kaviar database control.

Only 3 of 50 SNPs survived initial study criterion of at least twice as prevalent in patients, EFCAB4B involving calcium ion channel, LINC01171, and MORN2 genes.

We conclude the reported top-50 deleterious polymorphisms for Chronic Fatigue Syndrome were more likely the top-50 errors in the 23andMe and Kaviar databases.

In general, however, correlation of 23andMe control with ALFA was a respectable 0.93, suggesting an overall usefulness of 23andMe results for research purposes but only if caution is taken with chips and SNPs.
Click to expand...

https://www.medrxiv.org/content/10.1101/2020.10.27.20220939v1

Threads on genetic risks listed here:
https://www.s4me.info/threads/evide...n-me-cfs-discussion-thread.16117/#post-277072
 
Last edited by a moderator:
Good to see corrections of bad research - @wigglethemouse already noted many issues with the paper on the S4ME thread: https://www.s4me.info/threads/genet...erez-nathanson-klimas-et-al.9415/#post-171044
Two items from the Perez et al. results were immediate red flags. They showed the top 50 deleterious 39 snps with the greatest difference in frequency between CFS patients and control data. At the very top 40 of the leaderboard was a snp on the gene GPBAR1 that was 129,000 times more prevalent in CFS 41 patients. If this finding is accurate then the authors may well have discovered THE genetic cause of 42 CFS rather than just a predisposition.
Click to expand...

The Perez et al. discussion goes through, spelling out one by one, the function of each of the genes 142 from the top 10 in the table by summarizing what is known and including speculation for how these 143 factors may tie into CFS. For example, they suggest that decreased metabolism of xenobiotics may 144 be relevant to multiple chemical sensitivity disorder which is in turn relevant for some CFS cases and 145 a gene that is downregulated by sleep deprivation which in turn is a factor in chronic fatigue states. 146 The present reanalysis finds at the very least that such discussion and speculation are premature as 147 there is no evidence that any of those genes are relevant.
Click to expand...
 
This is a response to the study by Nancy Klimas which utilized 23andME data to try and identify genes associated with ME/CFS. 23andME is a commercial genetic testing company.

According to Bedford et al, the 23andME data is highly unreliable for the purpose of identifying genes associated with ME/CFS. I don't find this surprising because they only charge $200 for a report.
 
Last edited:
We conclude the reported top-50 deleterious polymorphisms for Chronic Fatigue Syndrome were more likely the top-50 errors in the 23andMe and Kaviar databases.
Click to expand...
Plain talking!

note that this new analysis has not yet been peer reviewed, but I am delighted to see healthy scrutiny of any ME/CFS research. It is all too rare. Note that the recent genetics review from Dibble, McGrath and Ponting also raised concerns about the Pérez study.
 
Michiel Tack said:
Good to see corrections of bad research - @wigglethemouse already noted many issues with the paper on the S4ME thread: https://www.s4me.info/threads/genet...erez-nathanson-klimas-et-al.9415/#post-171044
Click to expand...
Thanks for the mention. I did email one of the authors and never received a response. I tried to inform Solve ME/CFS about the issues before one of the authors presented a talk on their platform and again no response. Same with messages on Twitter and Facebook. Very disappointing as the errors were blatant and easy to find (also found by multiple users on PR). Because of this I no longer look at the NSU INIM teams papers in the positive light I once did.

Unfortunately it seems not acknowledging and not correcting bad Science papers is the norm in Science publishing from what I gather by following Elizabeth Bik on Twittter.
 
I know I've been critical of the Klimas/Perez study but here are a few good things about it.

1. They published the data in an Excel file in the supplementary section which allowed others to look at the work and check it. I wish more people did that.

2. It's allowed several other papers to cite it when looking at the quality of genetic studies in ME/CFS (3 of the top of my head). We now better understand the pitfalls with genetic data and this will help future genetic studies to be better quality.

3. Others looking to use 23andMe data have clear guidance on what to look out for with the paper of this thread title.

The real issue here is lack of funding in ME/CFS leading to shortcuts to make the best of the situation. Thankfully DecodeME seems to be reasonably funded and leverage well controlled quality practices already established at the UK Biobank team.
 
Simon M said:
Plain talking!

note that this new analysis has not yet been peer reviewed, but I am delighted to see healthy scrutiny of any ME/CFS research. It is all too rare. Note that the recent genetics review from Dibble, McGrath and Ponting also raised concerns about the Pérez study.
Click to expand...

One of the authors of the paper shared in this thread here! Thanks so much for the feedback. Indeed, the pointer at not being peer-reviewed yet is important. We submitted our own manuscript to Frontiers in Pediatrics (the same journal as the original Pérez study) end of July. Since then we haven't even been assigned any potential reviewers as Frontiers doesn't seem to have the fastest turn around ;-)
 
Trish said:
Hi @BastianGreshakeTzovaras, welcome to the forum. I can only imagine how frustrating it must be to have a paper waiting for months even to get to peer review. Thank you for your work. I wonder whether you have any more research planned on ME/CFS.
Click to expand...

Thanks @Trish! It's indeed very frustrating to wait that long to even get reviewers assigned and not even knowing when that might happen! As my expertise isn't really on ME/CFS as such but more on navigating 23andMe data I don't have any further work planned at this stage, but it's never say never. :-)
 
You must log in or register to reply here.
Back
Top Bottom