Overview of ME/CFS genes in DecodeME

Human Protein Atlas page for SRSF11 brain expression and it shows low tissue specificity, so its everywhere. Also widely expressed outside the brain. Apparently involved in mRNA processing and splicing but it’s not clear amd the data is quite old.
This gene encodes 54-kD nuclear protein that contains an arginine/serine-rich region similar to segments found in pre-mRNA splicing factors. Although the function of this protein is not yet known, structure and immunolocalization data suggest that it may play a role in pre-mRNA processing. Alternative splicing results in multiple transcript variants encoding different proteins. In addition, a pseudogene of this gene has been found on chromosome 12.[provided by RefSeq, Sep 2010
I’d certainly be interested in understanding more about why AG2 picked this up and what the wider analysis involved.
 
So I have results from AG2 but I do not know if it is prudent to post them as they are the product of an AI generated colab code which utilises AG2 API .

One thing to comment, a gene called SRSF11 -which came up to the analysis- and its effect to GABAergic neurons which has not been discussed anywhere in the forum (to the best of my knowledge).

If anyone else's analysis -who has the necessary background- identifies this gene, I could share the rest of results.
Probably deserves its own thread with methods, git repo, and explanation! Sounds interesting
 
My question had been how and who has the expertise to do this. AI tools are definitely helpful and maybe they’d be helpful enough for a few of us to do this? I also wondered if we could get a professional research team interested as it seems like a great project for someone. What do @mariovitali and @ME/CFS Science Blog think? Maybe @tralfamadorian97 and @forestglip too?
At least from a cursory glance at the paper and blog post, AlphaGenome 2 does sound like it has potential. Right now I am focused on the proteomics GWAS work. If my contribution would be helpful, I should be able to try out AG2 when I am done with proteomics.
 
At least from a cursory glance at the paper and blog post, AlphaGenome 2 does sound like it has potential. Right now I am focused on the proteomics GWAS work. If my contribution would be helpful, I should be able to try out AG2 when I am done with proteomics.
Is there any way we can support your work? You and @ME/CFS Science Blog work so hard and I would like to show my appreciation if I can.
 
Some first observations that stand out:

- The overrepresentation of brain and neuronal genes is confirmed in these top loci. So it isn't only apparent in tissue enrichment. Despite my best attempts, most of the immune loci remain ambiguous, such as the DARS2 locus at 1: 173,846,152. I suspect there's a tendency for immune-related genes to be in packed regions, while neurological genes are more often isolated and thus easier to identify.

- Unfortunately, there isn't one brain region or pathway that stands out. ARFGEF2, PLCL1, NEGR1, SOX6, VRK2, have GABA-related functions, while SHISA6, CACNA1E, and LRRC7 point to glutamate output. Some genes point to the hypothalamus (POU3F2), others to the striatum (PCDH17), and cerebellum (CA10, UNC13C, etc.).

- The strongest pattern is for setting the gain of synapses by bringing certain proteins to the surface. Most are neuronal genes, but there are no direct ones for ion channels, neurotransmitter receptors, or enzyme synthesis or degradation. Even vesicle release doesn't come up, except for the priming factor UNC13C. The ME/CFS genes are those that change the 'thermostat': the meaning and strength of certain synaptic connections. This is supported by the SynGO analysis, where "synapse adhesion between pre- and post-synapse", "regulation of postsynaptic density assembly" and "postsynaptic density membrane" were the categories most associated with ME/CFS. I suspect that this is how neuronal networks store and share information, including the 'symptom signal' that we speculated about before. It would explain why we don't see cell death or brain degeneration over time, as in, e.g., Parkinson's or Alzheimer's, and why the biggest correlations were with diseases such as chronic pain, depression, schizophrenia, and PTSD that probably involve setting synaptic gain in other networks.

- I did find one speculative connection between the genes. Both CYDL and ZNF644 are directly involved in REST and the G9a/GLP complex that keeps brain genes switched off in non-neuronal cells and turns them down to the right level in neurons. Other genes such as SUDS3/SRRM4, HTT, and MLLT10 also have connections to this system. I was wondering if this might be part of the explanation of why the symptom signal doesn't reset after a trigger such as an infection. It could also be a coincidental pattern though, as many of these genes have weak support and the gene category for REST didn't show enrichment in ME/CFS.
 
Last edited:
Unfortunately, there isn't one brain region or pathway that stands out. ARFGEF2, PLCL1, NEGR1, SOX6, VRK2, have GABA-related functions, while SHISA6, CACNA1E, and LRRC7 point to glutamate output.
This might be a dumb question. I was under the impression GABA and glutamate are kind of two sides of the same coin, or things that have to be in balance with each other. Is that not the case? I was wondering if those findings do in fact fit together.
 
Despite my best attempts, most of the immune loci remain ambiguous, such as the DARS2 locus at 1: 173,846,152. I suspect there's a tendency for immune-related genes to be in packed regions, while neurological genes are more often isolated and thus easier to identify.
Thanks for the effort here. Interesting observation. Hopefully SequenceME and other related projects will bring more clarity to these densely packed loci.

Which other loci would you say are immune or possibly so?

A lot of neurological leads and markers here but frustrating they dont yet coalesce into a more cohesive story. Seems like we're making progress though.
 
Back
Top Bottom