If I stop addressing my issues via the SNP reports, then I am back to the drs who think it's all in my head and keep worsening. So I will rather stick to the positive results I am having for both my husband and myself. I would be lost about what to do for my husband if it weren't the SNP reports. I have been hitting jackpot by sticking to those recommendations!
"Inventing random connections" and "no treatment at all" aren't the only options. The third option is to try understand the actual science instead of accusing innocent SNPs
I'm going to try to explain this, but I'm not much at good at it, so please disagree or question as desired:
- Human DNA contains in excess of 3 billion SNPs, spread out over 23 chromosomes. This is called the human genome.
- There are an estimated 20,000-25,000 genes on the human genome. Each gene is a sequence of SNPs, which work together to describe how a protein in the body should be made.
- Different parts of a gene do different things. The SNPs in some parts of a gene are translated into amino acids which form proteins. These coding parts are called exons.
- In the exons, each group of three consecutive SNPs spell out which amino acid should be created next on the protein which a gene creates. Thus CTC in an exon would indicate that a Leucine should be added.
- But even in exons, our bodies aren't too picky about perfect "spelling". Thus anything in an exon starting with CT and having a third SNP will add Leucine: CTA, CTC, CTG, and CTT all add Leucine, in addition to TTA and TTG. Hence many variations ("typos") are 100% known to be irrelevant and completely harmless.
- When a variation in an exon results in a different amino acid being created, it's called a missense mutation. For example, replacing CTC with CGC would result in Arginine being substituted for Leucine.
- Some missense mutations have little or no impact, due to the physical properties of the mutant amino acid being nearly identical to the normal amino acid. For example, if we have ATC instead of the normal CTC, Leucine is replaced by Isoleucine, which has identical physical properties to Leucine. Unless the exact structure of the Leucine in that exact position is needed for the protein to connect to another protein (which is unlikely) this change has no impact.
- Other missense mutations can have a big impact, especially if the new amino acid behaves very differently. Hence a protein might end up with a weaker bond due to the new amino acid, and perhaps it will break down at a lower temperature now, and not survive long enough to do its job. Or it might form too strong of a bond, and not break down when it should.
- Nonsense mutations result when a variation in a SNP results in a premature "stop" command, saying that the protein is finished. This is usually pretty bad news, unless it happens near the end of the exons, and not many amino acids are missing. But usually it results in the protein created by the gene being completely ineffective.
- Pathogenic missense and nonsense mutations which result in a gene becoming completely non-functional isn't always a problem, however - some genes are essentially duplicates of each other, or perform non-essential functions.
- Common missense mutations which have a large impact are not really pathogenic. They simply represent a normal, non-harmful variation. Hence if production of something is substantially down-regulated by one gene, perhaps its degradation by another gene is also down-regulated. Alternatively, the body might easily accommodate an excess of something, or just eliminate it faster if levels get higher.
- Things are a little trickier in the areas on genes which aren't in exons, known as introns. SNPs in the introns can do a variety of things.
- One function of intron SNPs is that repetitions of long sequences can identify where a gene starts and ends.
- Another function is that SNPs in these introns say where the exons are and that they should be translated into proteins.
- But mostly, we don't really know exactly what's happening in much of the introns, or why.
- Nonetheless, it's pretty uncommon for variations in the SNPs in introns to have a significant impact on gene function: at least 85% of such SNPs are on exons, not introns, even though exons are much much smaller than introns.
- Next we have huge stretches of SNPs between the genes, called intergenic. Some of the SNPs in intergenic bits can have an impact on nearby gene function, but most of it isn't known to do anything.
- On the ends of chromosomes we have telomeres, which are sacrificial lambs and don't do anything except provide a protective buffer for the SNPs further down which do do stuff. These telomeres get shorter as we age, at least partially in reaction to oxidative stress.
- But generally speaking, these intergenic SNPs and SNPs on telomeres aren't impacting genes.
So we can do a few things to determine which SNPs are actually causing issues. First we can ignore the intergenic and telomere SNPs. While there's always a chance that something will be discovered to show an impact, the current state of science says we don't know enough to draw any conclusions at all from those SNPs, and that they're unlikely to be relevant. There's not even a hint of how most of them might potentially have any impact.
Next, we can largely ignore SNPs on introns. Some have an impact, but thus far I've never seen any research showing them to have more than a small impact. It might be a statistically significant effect, but based on the effect sizes found in research these are usually up- or down-regulating a gene by 5% or less. It's small beans - they might add up, but one or two isn't doing much, and might even be completely offset by other mutations on the same gene: +5% from one, and -5% from another?
And then we can take a look at SNPs in exons, which is where nearly all of the real action is happening. Some of these SNPs in exons don't result in any change at all, but most do result in a change. If there's research, we can read it, and it will usually exactly how much gene function is reduced as a result of these mutations. Hence we know that MTHFR C677T +/- reduces MTHFR function to about 65% of normal, and is associated with an increase of birth defects if the mother doesn't compensate for it. Then we can compensate for it ourselves by eating more vegetables or taking a small dose of methylfolate, especially during pregnancy.
If there isn't research yet for missense mutations in exons, we might be able to predict the impact based on various highly complicated models, or by making simpler comparisons between the alternative amino acids.
But there are a few things we can't do. We can't assume that a lot of ME patients having a mutation is relevant to developing or perpetuating ME/SEID. For example, in the 31 ME patients I have data for, the average reduction in MTHFR function across all MTHFR SNPs tested by 23andMe is -31%. But for the 31 controls, it's an average of -32%. Additionally, no published research has demonstrated an excess of genetic or functional deficiency in MTHFR or active folates in ME/SEID patients. So while we can see where our functioning isn't optimal, by reading the research, we can also see that this dysfunction is pretty typical, again by reading the research.
Another thing we absolutely cannot ever do is assume that every SNP on a gene is relevant. There is a ton of data proving that a great many of those SNPs have no impact - as far as anyone can tell thus far, they are harmless background noise. There is also research showing that some are relevant, and that is what should be used to determine which SNPs might be causing us problems. Otherwise we are reduced to guessing.
And with 25,000 genes, and 3 billion SNPs, it is simply impossible for anyone to guess intelligently, or to draw any conclusions for treatments based on those guesses. If we take the absurd position of assuming that every SNP has a risky or +/+ version, then the result is that every human being has millions of "bad" SNPs. This is completely illogical, and an inexcusable abuse of the actual science or anything resembling statistics.
Thus we can look at our SNPs, see where they are on genes, and see if there's research about how those specific SNPs impact gene function. If they do impact function, we can determine if that can happen in a harmful manner, again by reading the research, or use it to direct further lab testing if it could be having a pathogenic impact.
Unless funds are unlimited, it simply isn't feasible to test every "treatment" for every SNP variation on every gene. And if taking that approach, there's no need to take SNPs into account in the first place, since that approach necessarily results in every gene being presumed broken without any actual supporting evidence. In which case, the SNPs themselves become completely irrelevant to the treatments being tried.
Hence if someone does want to use SNPs to find problems, they need to be used scientifically based upon the existing research, and not randomly. This shot-gun approach is actually undermining and even mocking the use of SNPs in finding actual problems. If we're going to use them, with or without the support of our doctors, we should be doing it correctly. Especially if doing it incorrectly is going to result in even less cooperation from our doctors.
And finally, we cannot make arbitrary connections between SNPs and effective treatments. Making a connection with any reliability requires using hundreds or thousands of patients. Otherwise we are just guessing that this one SNP means that this one treatment helps, even though there are billions of SNPs and thousands of potential treatments. Why that SNP and not the one next to it? Why not one on a completely different gene? Statistically we simply cannot make that connection reliably. This doesn't mean that a treatment doesn't or won't help, but it's important to not choose a random SNP to connect to a treatment - people might be overconfident that a treatment might help them, or people might think they shouldn't try a potentially helpful treatment because they don't have the "right" SNP. We have to be very critical and honest about what we actually know, and what we merely suspect.
