23andme - what's what

[TomBAvoider]

23andme announced changes as a result of their FDA contretemps. In their transition they need you to have had your info taken with one of the newer chips, sometime after late 2013. However, I personally have signed up with them and sent my sample in early 2010, so I don't qualify. They have not specified to me what happens now - do I re-submit another sample to be analyzed using their latest chips, or what. I am certainly happy to send in another sample, and even pay whatever the fee will be, but I'm not clear on what the process is, because they have indicated that they are using a staggered method of transitioning people in waves, and I clearly am not one of the early ones they've transitioned.

 

So, my question is: for those of you who are in a similar situation, where you've signed up with 23andme before the cutoff date of late 2013, and who have already been transitioned to the new system, could you describe the process? Also, while I did not take all the info as gospel, I did like how they highlighted risk areas (again, being fully aware that this is extremely speculative and there are massive interactions that are certainly not being accounted for when they identify a vulnerability) - is this going to be eliminated going forward, or do we retain the old information, or is it updated or what? Because if is being eliminated altogether, I'd like to archive it. FWIW, for me personally, the greatest benefit has been access to the raw data of my genome which I downloaded and put through the $5 promethease for an up to date report every few months. Of course, I have repeatedly come across snps referenced by various studies on pubmed where the data is simply not present on the 23andme raw data, I figure probably at least partially due to their having used an old chip.

 

Given that the raw data is 99% of my perceived benefit of 23andme (for myself), I'm wondering if I should even stick around. I mean, if I can get more comprehensive genome raw data elsewhere for a reasonable price, then I'm struggling to identify what 23andme is getting me - and if they completely eliminate, going forward, any risk assessments (for myself, not "being a carrier" for offspring I'll never have), then I just don't see the point. This transitioning might push me over to dropping 23andme altogether... unless someone here can share ways in which they use 23andme that have not occurred to me, that might be a reason to stick around. So my second question is: how do you use and benefit from 23andme?

[Dean Pomerleau]

Tom,

 

I got my 23andMe genetic test done long ago, well before 2010. I appreciated their old health reports, but I too have always taken them with a grain of salt. The most important thing I've learned from their report (and from running my data through Promethease) is a dramatically increased risk of macular degeneration, which I discussed here. These days, I mostly just use 23andMe to look up my genotype for specific SNPs that I come across in new research publications, using 23andMe's browse raw data search function. 

 

I haven't been told I need to give 23andMe a new sample to continue using their service. I'm curious, can you (or any other 23andMe subscribers), still see this Health Risk Report when you are logged in?  For me it looks like this (note: my two biggest elevated risks are eyesight related - macular degeneration 4.8x risk and glaucoma 2.9x risk):

 

 

--Dean

[TomBAvoider]

Yes, I can still see them. And I've got you beat! I have a virtual certainty of getting PC at close to 50%. I had a scare back in 2012 when my PSA came back as 7.59 ng/mL, even though the digital exam by the doc didn't detect any abnormalities; this combined with the 23andme risk assessment had me pretty convinced I was at serious risk, and I prepared myself for taking matters into my own hands should it come to that (I have zero desire to linger on in some hospital). However, before panicking too much, I read up and realized that there were several factors that may have drastically upped my PSA number (like the long bike ride I took the day before the PSA test etc.), so after a few weeks I got re-tested and came back with PSA at 0.94 and the more sensitive free PSA fraction at a pretty excellent 54%, which put my mind somewhat at ease. This year I got re-tested, and my PSA came back a bit higher, at 1.00 ng/mL, but still within norm, and DE showed no abnormalities. However, I am taking the risk of PC for myself very, very seriously, and have re-jiggered my supplement intake, changed aspects of my diet (used to blend black tea in with the green, now eliminated black altogether), upped exercise etc., all with a view that I want to do my very best to mitigate that particular risk. It's still shockingly high, but at least I feel like I can be prepared.

 

 

SEE NEW AND RECENTLY UPDATED REPORTS »

These reports provide information about your possible risk for developing certain health conditions based on genetics. Environmental and lifestyle factors also often play a large role in your risk for developing these conditions.

Elevated Risk

NAME CONFIDENCE YOUR RISK AVG. RISK COMPARED TO AVERAGE Prostate Cancer

42.0% 17.8% 2.36x

 

 

Colorectal Cancer

7.5% 5.6% 1.36x

 

 

Restless Legs Syndrome

2.5% 2.0% 1.25x

 

 

Crohn's Disease

2.1% 0.5% 4.05x

 

 

Ulcerative Colitis

1.00% 0.77% 1.30x

 

 

Esophageal Squamous Cell Carcinoma (ESCC)

0.43% 0.36% 1.21x

 

 

Celiac Disease

0.37% 0.12% 3.10x

 

 

Stomach Cancer (Gastric Cardia Adenocarcinoma)

0.28% 0.23% 1.22x

 

 

Primary Biliary Cirrhosis

0.14% 0.08% 1.72x

 

 

Abdominal Aortic Aneurysm

 

Edit: rats, I don't know what happened - it looked just like your formatting, Dean, when I pasted it in, but it came out here all wrong :(

[Dean Pomerleau]

TomB,

 

Thanks for sharing! I see your risk of prostate cancer is elevated, but FWIW, I've always been under the impression that virtually every man gets prostate cancer if they live long enough. Definitely worth keeping tabs on (as you're doing) but not a death sentence either.

 

Regarding formatting. I always use screen captured images, rather than relying on cut-and-paste of complex formatted data when I think formatting matters. The method I use is described here.

 

--Dean

 

[TomBAvoider]

Thank you, Dean! I'll try this as proof of concept on the "decreased risk" portion of my 23andme profile.

 

 

 

I guess this is slightly cheerier than the risk section, but equally unreliable. I noticed that my age related macular degeneration risk is opposed to yours FWIW, although one really wonders about all this, considering that when you go through promethease, you often have one variant indicated raised risk and another lowered, so who knows how you come out in the end once all of this is in play and interacting with the environment/lifestyle. The original idea behind the 23andme profile of risk was that they'd sift through all the conflicting data and arrive at the global risk profile - not that it was ever particularly realistic, given that we are still discovering additional information (which is why you have to keep running promethease every few months as more studies come out), and they have incomplete information to begin with (which is why they have to keep updating their chips). Bottom line, this is more entertainment than anything. I think the promethease way is more honest in that they simply list the + and the - as they come up, without trying to reach global conclusions.

 

Edit: hmm, still not as it should be - I used your method #3.

[Dean Pomerleau]

Wow Tom, FWIW, you're macular degeneration risk is 30x lower than mine! Lucky dog...

 

 

Edit: hmm, still not as it should be - I used your method #3.

 

Yes - method #3 is less than optimal. I use #2 to get nice embedded photos by uploading them to imgur.com. It is really quite straightforward.

 

--Dean

[TomBAvoider]

Today I received an offer from 23andme of $50 off a DNA kit, so I'd end up paying $149. And they also say that they are in the process of transitioning my account to the new kind, whatever that kind is, over the next few weeks, and that the experience of this new kit user will not be like my old experience and also I won't be able to "share" data. I'm still not clear on what this "new experience" means, and therefore in no position to evaluate this offer that expires by May 8. Just an amazing mess. I mean no disrespect to 23andme, but this whole thing is not exactly a model of transparency, PR and keeping the customer informed. Or maybe I'm just too dull to grasp it all. Oh well. As long as the new kit leads to more comprehensive data I can feed to promethease, I'm fine with it. Then again, if that's all I'm getting out of 23andme, maybe there's a better source of getting your genome sequenced? Has anyone looked into such an option?

[Dean Pomerleau]

Tom,

 

I got the same "Mother's Day" special offer for $49 off from 23andMe.

 

Is there evidence that the new chip they are using for sequencing covers a lot more genetic data that the older chip they used to sequence the genes of long-time customers like us? 

 

For me that is the only question that matters - since it sounds like any new "reports" they generate are going to be very sparse and watered down relative to what they used to give us (and which will apparently still be accessible to us old-timers after the 'transition' - just not much new in this regard).

 

Regarding alternative sequencing services see this and this thread for discussion of relatively inexpensive alternatives for getting full genome sequencing.

 

--Dean

[TomBAvoider]

Thanks, Dean. Alas, I don't know if the new 23andme kit uses a more advanced chip compared to what was used on my sample (back in early 2010), i.e. whether at the end of the process I'd end up with more complete raw data or not. My plan all along has been using promethease on the raw data from 23andme - but way too often when looking at various studies, I see that the SNP has not been sequenced by my old 23andme raw data set. And sad to say, the options you have so kindly pointed to, both have serious caveats. So for now, I'm in a holding pattern. I have been super sensitized to my prostate cancer risk, which looking at the promethease data makes it look like a virtual certainty... seriously, the sheer raft of studies that indicate crazy numbers (like 600% greater risk of contracting PC - which considering average risk is 17%, makes it something I can practically take to the cancer bank!) my risk level is off the charts. I'm amazed I'm not dead yet - and even more amazed that my PSA numbers are not worse (seriously, 1.00 at my age is definitely at the lower end of the normal range). I guess I'm grateful for the heads up, so that I can take heroic measures - sadly, there don't seem that many available... already in the course of my ordinary CRON lifestyle, the PC countermeasures suggested by the literature have been incorporated, which only leaves pharma interventions which I'm reluctant to get involved in (though I am thinking about aspirin preventatively). I certainly have upped my exercise and even resorted to speculatively beneficial supplements (provided there are legit studies supporting PC benefits, such as there are for grape seed extract - and so I take it, but there are no good ones for saw palmetto, and so I don't take that). But beyond more exercise and a raft of supplements, I'm already doing all I should be doing to minimize PC - I am however resigned to my fate, and I therefore avidly follow medical advances in curing PC. 

 

In any case, I just wanted to say that I do find a benefit to the gene sequencing. I am now alert for signs of trouble with PC, which I otherwise would have been totally oblivious to. I'm hoping for more and better sequencing, but I guess I'll have to wait to see what the developments are.

[TomBAvoider]

Another day, another offer of $50 discount on the 23andme kit. I've contacted them through their "submit a request", and if I hear back I'll share what was communicated. Meanwhile, out of pure curiosity, I wonder if anyone wants to chime in on the question of "how often do you run your 23andme data through Promethease?". I do it irregularly every few months, and being the impatient sort, I just pay the $5. I figure there is an avalanche of research papers out there constantly being generated which illuminate various alleles, and I wonder what the optimal "check in" frequency is. 

 

The other question/suggestion is, whether there exists some kind of app or interface that would check snpedia for research results that involve your raw data and issue alerts based around certain criteria (f.ex. "bad repute" results for your allele that has, say, Magnitude of >2). It seems shouldn't be super complicated to code, as you could keep your data private and keep polling snpedia for new info that matches any of your allele and automatically generates an alert to your specified criteria - that way you don't need to decide how often you need to run through your entire data, instead it continuously updates and alerts you to anything interesting. And not to put too great a strain on their servers, say, it could poll once a day, or even once a week. Is there such a thing?

 

FWIW, I had 24919 genotypes annotated - my biggest "bad reputes" are 1 report at mag. 3.5, 1 at 3.2, 12 at 3.0, 2 at 2.9, 1 at 2.7, 19 at 2.5, 2 at 2.4, 2 at 2.3, 4 at 2.2, 16 at 2.1, 97 at 2, 2 at 1.9, 7 at 1.8 and so on. The biggest one, the 3.5 is for rs1799807 I'm an AG "This single base substitution yields the 'atypical' version of BChE. This version has a lower affinity for many substrates, including many Cholinesterase Inhibitors (CIs)." As of late Nov. 2015.