Comment by aantix | Hacker News Reader

Comment by aantix | Hacker News Reader

I've posted this before, but it's worth repeating.

I recently discovered that I’m a slow metabolizer of caffeine. Via my raw 23andme data. https://you.23andme.com/tools/data/?query=rs762551

C/C genotype.

https://www.geneticlifehacks.com/liver-detox-genes-cyp1a2/#C...

Instead of coffee being a net positive, it actually takes caffeine 4X the time to process through my body. It also increases my cardiac event risk (most normal metabolizers it does not).

I can drink a cup of coffee at 8am and it still impacts my sleep the following night. My blood pressure remains really elevated (20+ points for systolic) the entire day.

All these decades I’ve been ingesting coffee thinking it’s a “net positive” - not for my genotype.

When the reality is I sleep like shit, my blood pressure is through the roof, I feel awful for a day or two of drinking a single cup, and it all makes sense now.

Even with just 30mg ingested this morning at 8am, I still feel very wide awake at 9pm.

When I totally detox from caffeine, go through the headaches, I ultimately feel calmer, and I get the best, deepest sleep.

Is there a way I can get this kind of analysis done and retain control of all the data that gets created? I have been interested in learning details about my genotype for a while, but not at the expense of having 23andMe keep a copy for whatever purpose they desire now or in the future.

Talk to your doctor. Genetic tests ordered via your physician may be covered by patient privacy laws in ways that 23andme customers aren’t.

Read the fine print though because you get what you pay for.

you will not get the richness of detail that 23andme provides though, since the results are mostly based off training data from other submissions to 23andme. Which is why it gets better over time; I was shown to be 20% Chinese, but now 4 years later my profile there reads that I'm completely Korean

Knowing your geographical ancestry - which, to me, seems slightly ridiculous, versus knowing some genetic predisposition for disease or problems like the one discussed above with caffeine or, say, ibuprofen, don't seem to be comparable in the slightest.

Their ancestry data is known to be quite dubious. They're really just guessing for the most part.

> I was shown to be 20% Chinese, but now 4 years later my profile there reads that I'm completely Korean

> you will not get the richness of detail that 23andme provides though

That richness you describe almost seems to be fabricated "vanity metrics" rather than anything useful

I work in human genetics and largely wonder the same thing. Short of sequencing your own genome, I'm not enthusiastic about most privacy policies.

To add context, I also am involved in public large-scale genetic studies (as a participant who has provided genetic data) such as All Of Us. There, I have no misapprehension that people will be studying my data.

I'm just saying that on the commercial side of things, I haven't found a satisfying consumer-facing way to pay for sequencing as a service (where I'm not trying to sequence at scale) without any strings attached.

The easiest test I can think of given GPs description is to take your blood pressure, abstain from coffee for a week (?) and then take a new blood pressure. It's not an RCT so it won't be definitive, but it may suggest to you whether further investigation is likely to pay off.

(Unless, of course, other confounders affect blood pressure unrelated to this effect. But for me personally, my blood pressure is usually fairly constant, at least compared to a 20 mmHg drop.

> The easiest test I can think of given GPs description is to take your blood pressure, abstain from coffee for a week (?) and then take a new blood pressure.

That entirely depends on how strong the caffeine dependency of OP is. For someone like me it would be anything but "easy" for first 2-3 days at least due to withdrawal symptoms (headache, poor concentration, tiredness).

Several months ago I saw a study claiming that decaf coffee prevents withdrawal symptoms, even if you know it's decaf. I was getting headaches pre-coffee in the morning, and tried it, and it worked so well I quit caffeinated coffee without any trouble. Just felt a little slow and sleepy the first week.

https://www.sydney.edu.au/news-opinion/news/2023/02/14/decaf...

The article you linked says it _reduced_, not _prevented_ the symptoms in people who knew it was a decaf.

Sorry, I misspoke, and that's my experience. I still had slight symptoms, but they were reduced by at least 90%. I'd found it pretty difficult to quit coffee before, but switching to decaf was effortless.

Yes, find a local sequencing provider and arrange to do a SNP chip or whole genome sequencing. In the contract ask that they delete your data after delivering it to you. This will be:

1. Expensive - probably at least 2 - 3 thousand dollars.

2. Require you to do your own analysis.

Obviously you can't be 100% sure they will delete your genomic data, but they have no incentive to keep it.

Don’t they have every incentive to keep it?

Such as? I’ve worked in genomics labs, they would be quite happy to delete stuff.

Why to delete anything sellable? I always do all kinds of illegal copies of somebody's PD on all places I ever worked as employee. Typically by stealing some papers from garbage because photographing takes too much time per one 'drop'. With enough wit of choosing workplaces I use to have some incomes from my strategy.

Jesus christ dude don't admit to crimes on a public forum

Any chance you're making a joke/point about someone else?

study it, sell it? Sadly, it seems privacy has gone out of fashion. I am happy to hear the genomics labs you've worked in would be happy to delete it.

There are thousands of sequenced human genomes available to access for research purposes (1000 genomes project, UK biobank etc) so one additional genome adds no marginal value.

They also have an incentive not to be sued for some astronomical figure if (more likely when) the data is found not to have been deleted.

Or find a European genetics laboratory that offers this service.

Even though you may not be European, the lab will need to follow the much stricter data protection policies of the EU and will probably not have a different handling for non EU customers (but ask them).

I write this as someone who has worked at one of those in the EU.

Since you are in the field, would you give some names of such labs in the EU? The quality of results when searching for "European genetics laboratory" aren't great.

> the lab will need to follow the much stricter data protection policies of the EU

But will they still retain it indefinitely on a server somewhere? If yes, it will leak eventually. Or the government will lobby to get access to it to "protect the children".

I think that they have to delete the data per your request to be GDPR compliant since no law requires keeping such date.

Re 2: do you know of tools that don't involve the cloud and allow you to do such analysis? Ideally FLOSS. I could only find DbSNP and SNPedia but they are datasets, not sure if there are tools built on top of that like Promethease.

Re WGS there are a lot of well established tool chains that are FLOSS (eg https://github.com/bcbio/bcbio-nextgen). You could run alignment and variant calling on a beefy workstation. A laptop would potentially work. Easy to test this with publicly available raw data. Another option: The sequencing provider often will run alignment and some default variant calling for you. Annotating and analysing these variants can be done on pretty much any computer, all with open source software. A SNP chip is even easier to deal with as the computational requirements are less.

Interpreting the results is a more manual process. Really depends on what you are interested in.

Related question: I have my genome but I find that the reports on polygenetic scores that you receive from [popular consumer WGS sequencing provider] are lacking accuracy. Is there a platform/ community of personal genetics enthusiast that crowdsources such analysis?

I haven't done a DNA test before, but I assume the identity verification is lax at best.

Would it be practical to just generate a "burner" persona to attach your test to? Yes your data technically will exist, but will be effectively anonymized

If anyone you're closely enough related to also gets sequenced, it will effectively de-anonymise you

I think Nebula makes the promise of anonymizing your data and permanently deleting it once the retention period has expired.

https://nebula.org/whole-genome-sequencing-dna-test/

Totally anecdotal, but have you tried coupling it with L-Theanine?

I've become extremely caffeine sensitive over the last few years -- anything more than about 80mg has me off the walls for the entire day. But ensuring I get L-Theanine along with it essentially blunts the response quite significantly and allows me to function normally.

why not just stop drinking caffeine, or drink less?

I did both. I take L-Theanine just in case since it's hard to approximate how much caffeine will be in a coffee.

Tea contains L-Theanine, you can simply switch if you still want the caffeine.

huh, that might explain why I've always found coffee "too much" for me, but never had significant problems with tea

While we're at it, here's an assessment of many drugs significantly affected by your CYP1a2 genotype: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1742-7843...

I'd recommend checking out your CYP2D6 too. Proper recommendations haven't yet reached the clinic, so if you want personalized medicine then you'll have to educate yourself. Caveat, for dozens of reasons, differing genotypes won't necessarily affect drug pharmacokinetics, even if the metabolic pathway suggests it should.

Liver enzymes are cool.

> I'd recommend checking out your CYP2D6 too.

I have the nonfunctional genotype: Some effects. I can't metabolize codeine into its active drug, so it's just a trash useless drug to me that was prescribed a few times in the past to no effect. Additionally, DXM cough meds stays with me a crazy long time so I have to increase dosage interval else I get high off it. No doubt affects other common medications.

There is a test: https://en.wikipedia.org/wiki/AmpliChip_CYP450_Test if anyone knows more the Wikipedia page needs updating. Maybe there's more tests now. Just seems crazy none of these tests seem to be routine yet to keep in patient file. This is not 'new' stuff anymore.

>Proper recommendations haven't yet reached the clinic

Completely agree. I've seen four cardiologists, two nephrologists in my years of dealing with high blood pressure.

They basically just mix and match different BP drugs, hoping something works.

It's out of the stone age.

I just checked, and I'm C/C too, but I don't have the same experience: I can drink a coffee after diner and still fall sleep easily.

It could be that it's more complex than a toggle on/off, and more like a polygenic trait.

Sleep quality might still be affected.

How old are you? Caffeine has affected me much more strongly as I’ve aged to the point where I just do freshly ground decaf now. Some of the methods to remove caffeine don’t completely ruin the flavour these days so if you love coffee it can be worth trying…

mid 30

It's not just about time-to-sleep, it's about sleep quality.

My only suggestion would be to fully go off of caffeine and rate your sleep quality off of caffeine.

I went years of drinking caffeine. So much so, that I had forgotten what deep sleep really felt like until finally removing caffeine from my daily routine.

It's a massive difference in sleep quality. With caffeine, the sleep is superficial. I get up to urinate a lot more. Lighter sounds can wake me up.

Without caffeine, I find myself sleeping so deeply, that my body stays in the same position, and I can wake up to tension in certain areas because I haven't moved for hours. The deep sleep feels amazing.

It's good to try different things. Try a month without caffeine, alcohol or other stimulants for a month to find a baseline. Then you can find what's good for you.

I have a friend psychiatrist who told that she had given the standard first question and advice for over decade for sleep problems: "How much you drink caffeinated beverages? Maybe you should drink less" before she took the advice herself.

She was amazed how well she slept without caffeine even when she didn't think she had sleeping problems or difficulties. She just didn't remember what it was to sleep really well anymore.

I haven't been tested but I read about this, and I think I may have something similar. One cup of cold-brew coffee and I'm good for the day. I don't start getting caffeine withdrawal symptoms (headache etc.) until mid-day the next day.

I used to drink caffeinated soda throughout the day, quitting around 5 hours before bedtime, and I had constant sleep issues. I stopped that around 10 years ago and switched to my one-cup-per-day system. Now I sleep fine.

Very interesting! I generally avoid coffee after the morning since I know I'm sensitive to it, but your link points to the specific gene I have that causes it. I wonder if I should reduce my caffeine intake even in the morning. Is there a good way to find out other relevant genes without uploading my 23andMe data to unknown websites?

Get your DNA sequenced with Nebula. They seem to be pretty committed to privacy.

https://nebula.org/whole-genome-sequencing-dna-test/

Another interesting side not to this.

Of the times that I have lost significant weight, both have been times when I quit caffeine completely.

I understand - it's hard to tease out whether it was the caffeine impeding the weight loss I get it.

It could be just as simple as I sleep better without caffeine and the better sleep quality improves hormonal balance, satiety, etc.

Same here. I've quit coffee about 15 times in the last 10 years.

EVERY. SINGLE. TIME, the result is the same: high blood pressure, bad sleep, constant lethargy (which can only be fixed by consuming more coffee), frequent headaches... all goes away after one week of detox.

Why would coffee be a net positive, even in general without any genetic mutations?

CGP Grey: Coffee: The Greatest Addiction Ever: https://www.youtube.com/watch?v=OTVE5iPMKLg

Hmmm maybe time for me to activate that health portion of my 23andme. I thought it was just silly stuff included but stuff like this would be interesting and useful.

Is that the rs4410790 marker ?

rs762551

Reply on news.ycombinator.com