WGS (whole genome sequencing) findings possibly linked to mitochondrial/metabolic dysfunction – anyone with similar results?

[deleder2k]

Hi everyone,

I’m a a male in my 30s with ME/CFS (diagnosed in 2012 using ICC criteria), now mostly housebound with severe post-exertional malaise (PEM), disabling muscle pain and cramps (especially in the thighs), and profound fatigue. After years of inconclusive testing, I recently did whole genome sequencing (WGS) and have been investigating potential genetic causes using tools like VEP, dbNSFP, Exomiser, and IGV.

Some of my most relevant genetic findings include:

ATAD3A p.Ala374Thr (REVEL 0.91)


Missense variant in a gene involved in mitochondrial dynamics, mtDNA organization, and cholesterol transport. High pathogenicity scores.

SLC25A5

Three heterozygous variants: two frameshift and one high-REVEL missense. This X-linked gene encodes the ADP/ATP carrier ANT2 in the mitochondrial inner membrane. Biallelic loss is embryonically lethal, but X-linked heterozygous disruption may have effects in males.

CPT1A p.Ala275Thr

A relatively common missense variant, but some studies link it to reduced fatty acid oxidation. It may act as a functional modifier under metabolic stress.


Other VUS in mitochondrial/metabolic genes:

ATAD3B, ATAD3C, ATPAF2, PCCA, OGDH, POLG2, TWNK, SLC25A4

These are all genes involved in mtDNA maintenance, TCA cycle, oxidative phosphorylation, or fatty acid metabolism.

My muscle biopsy (2019) showed mild fat infiltration and slightly elevated ammonia. CK and EMG were normal. I’ve had recurrent episodes of dark/bloody urine after exertion (likely hemoglobinuria, but no confirmed myoglobin).

Core symptoms:

• Severe thigh pain and cramping after minimal exertion (stairs, standing still, squatting).
• Post-exertional malaise lasting days to weeks.
• Consistent symptom relief from alcohol (high dose): increased energy, less pain and cramping. Effects peak after ~3–6 hours and gradually wear off over ~24 hours.
• Orthostatic intolerance / POTS.
• Brain fog, sound sensitivity, and very poor stress tolerance.

What I’m wondering:

1. Has anyone else done WGS/WES and found rare variants in similar genes (e.g. SLC25A5, ATAD3A, CPT1A, or other mitochondrial/metabolic genes)?
2. Has fatty acid oxidation or mitochondrial dysfunction ever been explored in your ME/CFS diagnosis?
3. Would anyone be open to comparing findings or sharing similar variant patterns?

I’m not a geneticist, but I’ve been learning a lot through tools like Ensembl, Exomiser, and manual review in IGV. I have both VCF and FASTA files if anyone else is digging into their own data and wants to collaborate.

I also came across a recent 2025 study by Hansen et al. (J Med Virol) which found rare heterozygous variants in mitochondrial genes (e.g., OXPHOS and fatty acid metabolism) in long COVID patients with ME/CFS-like symptoms. Their “multi-hit” hypothesis (multiple heterozygous variants combining to cause energy failure under stress) seems very relevant. Although my case isn’t post-COVID, the overlap is striking and lends support to a similar mechanism.

Thanks so much for reading — any thoughts, feedback, or similar experiences would be very welcome!

 

[wigglethemouse]

There is another tool you can use to analyse VCF files, written by someone with ME/CFS.

Upload Genome | GenVue from Genetic Genie

Some thoughts:
* Variant frequency is important to look at and understand. Focus on the lower frequency ones to start with.
* Clinvar is useful to see if the variant is known to cause disease.
* Gnomad is a useful database that runs variants through loss of function silicon assessment. It also lists the protein ID made as a consequence.
* Looking at CADD score can be helpful to determine the possible severity of the variant
* Finding papers related to the SNP that test for protein expression can be helpful, although hard to find that kind of info.
* Genecards can also be helpful. It has a disease relation section.
* Be sure to check allele count for quality. Sometimes interesting findings have lower quality reads.

The biggest issue I have is that for most variants I've found in my data, it's really hard to determine what the effect really is, especially compound heterozygous variants and heterozygous variants associated with autosominal dominant disease. Sometimes there is very little info about the gene itself. On top of that having a variant that causes disease doesn't necessarily mean that you have that disease problem. Our genetics have many backup pathways.

Good luck on your journey.

 

[deleder2k]

There is another tool you can use to analyse VCF files, written by someone with ME/CFS.

Upload Genome | GenVue from Genetic Genie

Some thoughts:
* Variant frequency is important to look at and understand. Focus on the lower frequency ones to start with.
* Clinvar is useful to see if the variant is known to cause disease.
* Gnomad is a useful database that runs variants through loss of function silicon assessment. It also lists the protein ID made as a consequence.
* Looking at CADD score can be helpful to determine the possible severity of the variant
* Finding papers related to the SNP that test for protein expression can be helpful, although hard to find that kind of info.
* Genecards can also be helpful. It has a disease relation section.
* Be sure to check allele count for quality. Sometimes interesting findings have lower quality reads.

The biggest issue I have is that for most variants I've found in my data, it's really hard to determine what the effect really is, especially compound heterozygous variants and heterozygous variants associated with autosominal dominant disease. Sometimes there is very little info about the gene itself. On top of that having a variant that causes disease doesn't necessarily mean that you have that disease problem. Our genetics have many backup pathways.

Good luck on your journey.

Thanks for sharing this — I completely agree, especially with how difficult it is to interpret compound heterozygosity or heterozygous variants in poorly understood genes. I’ve gotten a lot of help with ChatGPT/GenomeGPT.


What’s particularly intriguing is that I experience a dramatic (though temporary) improvement in muscle pain, fatigue, and physical stamina after alcohol — which makes me wonder about a possible mitochondrial component, maybe via NADH/NAD+ metabolism or mPTP regulation.





I’ve used tools like MutationDistiller, Ensembl VEP, and Exomiser to prioritize variants, but I’m curious how GenVue compares — have you found it more helpful than tools like Franklin (by Genoox)? You van upload your VCF to franklin. Select allell frequency (total i think, not internal) under 1%, and then for instance select CADD > 20. That will yield you interesting results. Export the file and upload it to chatgpt and have it analyze the CSV file. If you haven’t tell it that you have ME and your symptoms. I even uploaded 100 pages from my hospital. ChatGPT and Perplexity concludes that it is a 95-99% probability that I have a mitochrondial disease (nuclear, not in the MT DNA).

If you havent you can also have ChatGPT analyze your MtDNA if you have it.

 

[wigglethemouse]

What’s particularly intriguing is that I experience a dramatic (though temporary) improvement in muscle pain, fatigue, and physical stamina after alcohol — which makes me wonder about a possible mitochondrial component, maybe via NADH/NAD+ metabolism or mPTP regulation.

There seems to be a small subset that experience this at times. The majority report getting worse. I'd love to know why it helps.

Thanks for the heads up on the analysis. I've only really used gene.iobio phenoanalyzer and an older hg37 custom tool..

 

[jnmaciuch]

What’s particularly intriguing is that I experience a dramatic (though temporary) improvement in muscle pain, fatigue, and physical stamina after alcohol — which makes me wonder about a possible mitochondrial component, maybe via NADH/NAD+ metabolism or mPTP regulation.

Hi @deleder2k! One of my first research projects in ME/CFS was looking at alcohol intolerance--it seems to be much more common to experience a worsening of symptoms rather than an improvement with alcohol (like @wigglethemouse said). That being said, it's not the first time I've heard a story like yours.

My hypothesis was that the alcohol intolerance was due to an NAD/NADH balance issue (though I only had survey data so could not confirm this mechanistically). It's within the realm of possibility for redox balance issues to go in either direction, though, in which case alcohol might have the opposite effect.

Interestingly, I had personal experience of malic acid (a shuttle for H+ into the mitochondria) ameliorating my own alcohol intolerance. Good luck on your detective journey, I hope you get some useful information out of it!

 

[wigglethemouse]

Consistent symptom relief from alcohol (high dose): increased energy, less pain and cramping. Effects peak after ~3–6 hours and gradually wear off over ~24 hours.

A friend came across this paper a few years ago.
Ethanol metabolism: The good, the bad, and the ugly

We found this section interesting.
"Evidence supports the view that, following ethanol ingestion, elevated blood acetate becomes a major energy source for resting and moderately active muscle, consistent with associated suppression of AMPK activity..."

 

[jnmaciuch]

A friend came across this paper a few years ago.
Ethanol metabolism: The good, the bad, and the ugly

We found this section interesting.
"Evidence supports the view that, following ethanol ingestion, elevated blood acetate becomes a major energy source for resting and moderately active muscle, consistent with associated suppression of AMPK activity..."

That's quite interesting, I suppose it makes sense that increased supply of acetate would lead to increased Acetyl-CoA for the citric acid cycle. Julia Newton's group found decreased AMPK activation in ME/CFS, so that's another possibility of why alcohol tends to make most pwME much worse. But there's room for lots of variation on the exact metabolic details between pwME.

 

[Hoopoe]

I too have a suspicious variant in SLC25A5.

What I learned from analyzing my own genome is that we all have tens of genetic variants that will be considered pathogenic by variant effect prediction tools. So it's not difficult to find variants that look like they might have something to do with one's symptoms. The hard part is figuring out which ones, if any, have something to do with it.

Some genes tend to have one or several of these potentially pathogenic variants. That probably means they are regions of the DNA where mutations occur frequently and are also more easily tolerated, so that the usual variant effect prediction tools aren't particularly useful.

In many genes the body can tolerate the functional loss of one allele without significant consequences. For example, I also have a pathogenic variant in ACADM gene which is probably reducing the enzymatic efficiency of the protein by about half. The protein is required to utilize certain fatty acids and a loss of enzyme function can lead to a serious metabolic disease. I have wondered whether this could explain my tendence to hypoglycemia and poor fasting tolerance (which are key symptoms of the associated metabolic disease), but based on what is known, it appears one would not expect significant negative consequences from a single pathogenic variant.

I also have a pathogenic ACE variant which came up in a recent paper on genetics of ME/CFS.

It's important to distinguish between pathogenic variants that might be contributing a little to the disease, and pathogenic variants with such a large negative effect that they alone are sufficient to cause disease.

 

[wigglethemouse]

gene.iobio highlights 4 moderate to high impact heterozygous rare variants I have on gene SLC25A5 but Genvue doesn't list them (must filter out for some reason?). I find having 4 a bit suspicious. They are on the X chromosome and I only have 1 X.

EDIT : I checked one of the genes on gnomAD. Looks like this location on the chromosome may have some known quality issues. Probably why Genvue discarded it as it checks the databases for things like this.

This variant is covered in fewer than 50% of individuals in gnomAD v3.1.2 genomes. This may indicate a low-quality site.

3 out of the 4 genes have quality issues shown on gnomAD for my variants.

 

[deleder2k]

I too have a suspicious variant in SLC25A5.

What I learned from analyzing my own genome is that we all have tens of genetic variants that will be considered pathogenic by variant effect prediction tools. So it's not difficult to find variants that look like they might have something to do with one's symptoms. The hard part is figuring out which ones, if any, have something to do with it.

Some genes tend to have one or several of these potentially pathogenic variants. That probably means they are regions of the DNA where mutations occur frequently and are also more easily tolerated, so that the usual variant effect prediction tools aren't particularly useful.

In many genes the body can tolerate the functional loss of one allele without significant consequences. For example, I also have a pathogenic variant in ACADM gene which is probably reducing the enzymatic efficiency of the protein by about half. The protein is required to utilize certain fatty acids and a loss of enzyme function can lead to a serious metabolic disease. I have wondered whether this could explain my tendence to hypoglycemia and poor fasting tolerance (which are key symptoms of the associated metabolic disease), but based on what is known, it appears one would not expect significant negative consequences from a single pathogenic variant.

I also have a pathogenic ACE variant which came up in a recent paper on genetics of ME/CFS.

It's important to distinguish between pathogenic variants that might be contributing a little to the disease, and pathogenic variants with such a large negative effect that they alone are sufficient to cause disease.

Have you tried a diet? People with CPT1/2 mutations and also different mutations in other genes need to follow a strict diet. MCT is often used.

Yes, I have some strange variants in SLC25A5, but one or two of them are normal for 4% of the population. But I do have some very rare variants too. I think they are most interesting.
It does seem that SLC25A5 is a difficult gene to map. I will try to do CNV analysis to figure out if the results are reliable.
These are my variants in the mean time:
My confirmed SLC25A5 variants (hg38):

1. chrX:119469998:CT>C (rs759019641)
   • HGVS: c.192delT → p.Phe64Leufs*8
   • Consequence: Frameshift variant, introduces premature stop codon → likely loss-of-function
   • gnomAD allele frequency: 0.001% (1 in 100,000)
   • REVEL score: 0.81 (highly pathogenic)
   • ClinVar: Not reported
   • Status: Confirmed as hemizygous and real in IGV
   ⮕ This is the most likely pathogenic variant.
2. chrX:119469743:A>AG (rs113356560)
   • HGVS: c.132+1_132+2insG
   • Consequence: Splice-region/intronic variant (may affect splicing)
   • gnomAD allele frequency: 0.001%
   • ClinVar: Uncertain significance
   • Status: Confirmed as hemizygous in IGV
3. chrX:119469901:G>A (rs199707714)
   • HGVS: c.147C>T → p.Ser49=
   • Consequence: Synonymous variant, but located near splice site (possible regulatory effect)
   • gnomAD allele frequency: 0.0015%
   • ClinVar: Uncertain significance
   • Status: Confirmed as hemizygous in IGV

All three variants are extremely rare, hemizygous (I’m male), and confirmed visually in IGV. One is a frameshift with high REVEL score, and together with the others (which may affect splicing), this supports possible X-linked loss of function in SLC25A5. This gene encodes a mitochondrial ADP/ATP translocase (ANT2), making it a strong candidate in my case with muscle pain, post-exertional malaise, and fatigue.

 

[wigglethemouse]

chrX:119469998:CT>C (rs759019641)

Two warnings on gnomAD.
1. Failed allel specific VSQR filter
2. Warning This variant is covered in fewer than 50% of individuals in gnomAD v3.1.2 genomes. This may indicate a low-quality site.

chrX:119469743:A>AG (rs113356560)

3 warnings on gnomAD. Same two as before + inbreeding coefficient warning.

chrX:119469901:G>A (rs199707714)

2 warnings on gnomAD similar to above. Low quality site.

Looks like you are in a very similar position to me and my 4 variants on SLC25A5.