23andme – yes, me – part 3

The email is titled “Your Genetic Profile is Ready at 23andMe!” It arrived on June 21, a shade under 5 weeks after sample arrival and well ahead of the estimated 6-8 weeks. This is what we’ve been waiting for.



Welcome to me!

On logging in to the 23andMe website, you’re presented with a news feed which confirms that your results are ready. To the left is a navigation menu divided into 4 main sections: My Health, My Ancestry, Sharing & Community and 23andWe. The first 3 are the most relevant with regard to your genomic data.

My Health
The Health menu has 5 sub-menus: Disease Risk, Carrier Status, Drug Response, Traits and Health Labs.
Disease Risk


Marker Effects

Clicking on “Disease Risk” presents 3 summary tables for elevated, decreased and typical risk, respectively, covering around 80 conditions. Each table shows the disease name, a confidence level from one to four stars (indicating to what degree the markers are believed to associate with the disease) and an indication of your risk, compared with an average.
Click the disease name and you are taken to a page with more detailed information. First, there is a summary of the disease, followed by your odds calculator, an estimate of genetic versus environmental contributions to disease risk, some health advice and a graphic that shows the estimated contribution of your genotyped marker(s) to your risk.

Here’s a table summarising my results, with confidence levels:

Risk **** *** ** *
Elevated 3 5 2 1
Decreased 6 8 6 1
Typical 14 18 15 2

I think the data presentation of disease risk is excellent. The information is summarised effectively – neither too brief nor too verbose. There are plenty of links to further sources of information, such as the studies on which the statistics are based. Most importantly, it is made very clear that disease likelihood is affected by both genetic and environmental factors. One of my stronger decreased risks, for example is for melanoma, but does that mean I can spend hours unprotected from the fierce Australian summer sun without concern? It certainly does not.

On a personal level, I’m pleased with my risk assessments. There are only 3 high-confidence elevated risks, the fold-elevation is pretty low in all cases and there are strong environmental factors too.

Carrier Status
The Carrier Status section lists 24 diseases for which you may be a carrier, given a certain genotype, all of which have a four-star confidence level. Clicking on the disease shows a summary, your genotype and more information about the gene(s) and how their variants can lead to disease. Once again, this information is presented very effectively.

I’m happy to report that I am “variant absent” for all of these markers. Which is not to say that I don’t carry any deleterious mutations – we may not yet know all markers for a condition – but this information was both useful and comforting to me.

Drug Response
Drug Response lists 17 common drugs and attempts to predict your response to them, based on genotype. Once again, these data come with a confidence level and further information about the marker(s). Not to mention disclaimers to remind you, rightly, that drug prescription is the domain of medical professionals.

Not much of note for me personally here; I am largely “typical”, with the exception that my rs762551 is CC, making me a slow caffeine metaboliser. This surprised me as I can consume what I’d consider to be relatively high amounts of caffeine without feeling “jumpy”.

Traits are good fun and perhaps one of the first things we learn in genetics at school (tongue-rolling, bitter tasting and so on). In the case of 23andMe, they are also reassuring since they’re a good indication that you received the correct data. Yes, I do have blue eyes and wet earwax.
Once again, traits are presented in a list (34 in total) with confidence levels and a detailed report for each trait. This is an area where I found the low-confidence traits quite unconvincing, but many of the higher confidence predictions to be spot on.

Personal observations of interest: CC in rs1815739 gives me two working copies of the “fast-twitch” muscle gene, more common to sprinters but also found in endurance athletes. I’m hardly an athlete of any kind right now but as a fit young man, I was much more a cross-country enthusiast than a sprinter. AG in rs10195871 means that I may still be producing fetal haemoglobin. I’m not sure what to make of that, except that it’s intriguing and I must say, a little bizarre, for some reason. I guess it’s something that had never crossed my mind.

Health Labs
Here, you’ll find some experimental tools for playing with your data, which I won’t cover in this post.

My Ancestry
Moving to ancestry, we’re presented with 6 sub-menus: Maternal Line, Paternal Line, Relative Finder, Ancestry Painting, Global Similarity and Ancestry Labs.

Maternal Line
Your maternal haplogroup is determined by your mitochondrial DNA (which is passed down the female line). 23andMe provide geographical maps, a history, a tree and links to further information to illustrate and describe haplogroups.
Mine is H1, which is very common in Western Europe and probably originated on the Iberian peninsula, where a woman some 13 000 years ago was enjoying the warmer climate, waiting for the northern ice sheet to retreat. Eventually, one of her descendants settled near what came to be Manchester and gave rise to my mother’s side of the family.

Paternal Line
If you’re a boy, your Y chromosome, inherited through the male line, determines your paternal haplogroup. Mine is I2b1 which again, is common in north-west Europe and advanced north as the ice retreated. The copy that I inherited got as far as Ruislip, which my dad will tell you is still in Middlesex, not London. From there it travelled up the motorway, collecting mum’s H1 en route and combining with it here. Welcome to me, indeed.

If you’re interested in haplogroups, Wikipedia has several good introductory articles.

Relative Finder
I won’t cover the Relative Finder in any detail here, since I share at most 0.37% DNA with other 23andMe users and don’t expect, at this stage, to discover any close relatives in their database. However, I believe that others have done so and it looks like a fun way to use your genotype data – provided there are no unpleasant surprises! In fact, you’ll see a warning to this effect the first time that you visit the link.

Ancestry Painting


100% white boy from NW Europe

Ancestry painting displays a graphic representation of chromosomes 1-22 and colours segments according to broad geographic origin: African, Asian or European. Unsurprisingly, given my haplotypes and family history, I’m 100% European. This was a tremendous disappointment to my Australian-Malaysian partner, who likes to believe that there’s a little bit of Asian in everyone :-)

Global Similarity


The Northern Europeans

Just to confirm your ancestry beyond doubt, the Global Similarity view shows a geographical map, a bar chart and an advanced view of your genetic similarity to human sub-populations.
There I am, smack in the middle of the Northern Europeans.

Ancestry Labs
Like Health Labs, Ancestry Labs is a sandbox with experimental features for looking at your ancestry data.

Sharing & Community
Under this menu, you’ll find features that allow you to compare your genotype with various groups and individuals, such as family members or friends with whom you connect at 23andMe. This includes whole genotype comparison or a selection of particular traits. I have not explored or used these features very much so far.

Browse/Download Raw Data


SNPs in my PRAC

This is obviously the most exciting section for the bioinformaticians in the audience. Under “Account” at the top of the screen is a “Browse Raw Data” link. This provides two options. The first is a very nice graphical view of each chromosome. You can click a chromosome to see its SNPs and genotypes or search using a specific gene (an HGNC symbol such as PRAC works nicely).
There’s also a “download raw data” link. This generates a zipped file of around 5 MB (around 15 MB when unzipped). It’s plain text with comments in the header, followed by 4 columns and 578 319 lines for SNP ID, chromosome, position and genotype. Coordinates are based on the NCBI36/HG18 genome build. A few sample lines shown below:

# rsid  chromosome      position        genotype
rs3094315       1       742429  AA
rs12562034      1       758311  GG
rs3934834       1       995669  CC
rs9442372       1       1008567 GG
rs3737728       1       1011278 GG
rs11260588      1       1011521 GG
rs6687776       1       1020428 CC
rs9651273       1       1021403 GG
rs4970405       1       1038818 AA
rs12726255      1       1039813 AA

You can imagine the possibilities provided by these data. Some ideas: a GBrowse installation, displaying your own SNPs with links to relevant databases. Calculate the amino acid substitutions arising from SNPs in protein-coding regions. Homology modelling of said proteins. Predict alterations in transcription factor binding affinity for SNPs in regulatory regions. Contribute your (anonymised) data to a database for improved GWA studies.
There’s a little more information at SNPedia, which also points out that deletions and insertions are encoded in the file, using “D” or “I”.

Summary – 23andMe
I’ve had an entirely positive experience with 23andMe, from start to finish. Ordering and sample collection was handled very professionally, notifications were prompt and informative and I’m really impressed with their website and the way in which complex data are presented. From what I’ve heard, they also deal quickly and openly with problems when they (inevitably) arise.
I’ll admit that their normal prices are higher than what I’d be prepared to pay: not because they’re unjustified given the service but because for me, this is certainly a “luxury item” (spending on non-necessities always makes me uncomfortable; it’s my tough northern upbringing – and perhaps genes!) However, I was more than happy with the special DNA Day offer – may there be many more.

Summary – Me
Getting genotyped has been an interesting and rewarding experience. It’s also ongoing, since the 23andMe website is constantly updated with new data, tools and findings.
Assuming that my results are accurate, they confirm what I’ve long suspected: genetically, I’m a very fortunate individual, with very few causes for concern. What I find interesting is that my genetic data have made me think more about environment and lifestyle than genetics.

In my late teens and early twenties, I was a fit and healthy young man who swam, ran, cycled and walked everywhere. I’d think nothing of stomping 40 km across the Scottish Highlands in mid-winter, taking in a half-dozen 1100 m peaks along the way. Last month, I visited a doctor with a medical complaint for the first time in over 20 years.

In my later adult life, I have not been so kind to myself. This ranges from a more sedentary lifestyle, to downright abuse in times of psychological stress. I know now that my poor decisions are far more likely to “get me” than my genes and that, in a sense, I’m abusing my genetic heritage if I don’t look after myself. I also have a couple of genetic pointers to particular conditions that I should keep in mind, as I get older.

Perhaps that would be my message to anyone considering a genotyping service. Your genes are interesting, but it’s the decisions you make that shape your life. Be good to yourself.

23 thoughts on “23andme – yes, me – part 3

    • Difficult question. I know very little about genetic counselling. There are links and information about it scattered throughout the 23andMe website. I guess the data returned falls into different categories regarding “concern”: traits are pretty innocuous, disease associations are not cut and dried, but finding out that you are a carrier could be upsetting.

      I guess I’d advise people to understand that having a particular genotype does not mean “you will get this disease” and to be aware of environmental contributions. And perhaps, to stay away if they would “rather not know”.

    • I was not too concerned. Partly because I know that my general level of health is good, partly because I’m a pretty relaxed individual, most of the time. Some people did express concern of the “but what if you find out something bad?” variety. To be honest, I don’t know why I gave it so little thought. Is there a SNP for “failure to imagine consequences” ?

  1. Hi Neil! Loved your summary — quite refreshing after some of the recent drama in the industry ;). And the excitement of learning about oneself and the feeling of opening wide the doors of possibility was tangible. Glad you found the information well-presented as well — though I think it is still quite complex for most people. It’s always a challenge!

    • Yes, I guess I view the information as an “insider”, as opposed to someone without a lot of biological education. I do think they’ve done a pretty good job though, particularly in separating the key points from the extras.

  2. So, I guess you are not genetically a woman — which, if I recall from a previous posting, would have been grounds for refusing to discuss the results publicly.

    But now you’ve encouraged me to check out 23 and me — no matter that logic tells me that if I wait 5 years, this sort of thing will be worth $20 at most.

  3. Interesting post, Neil. As a researcher who works with a lot of genetic counselors, I’d respond to Duncan’s comment by saying, yes.

    It’s one thing to face certain possibilities – rather, probabilities – with equanimity when they are abstract. It’s another when it has possible – or probable? – outcomes for oneself and one’s children. We just aren’t rational this way. Having an expert outsider help guide you through this is probably a good co-investment if you go as far as personal genotyping.

    Thanks also, for the plug to make anonymous donation of the results. I support the idea in theory, but the WTCCC for instance doesn’t accept such donations, do you know who does? Getting access to such data is rightfully controlled, as well (cf http://ctsi.ucsf.edu/blogs/ethics/genomic-sequencing-identifying-anonymous).

  4. Thanks for providing such a helpful and considered report on your results. There is in fact already quite a lot you can do with your raw data from 23andMe. You can upload your Y-chromosome results to Adriana Squecco’s spreadsheet. You can find full instructions at the bottom of this webpage:


    A number of new SNPs (single-nucleotide polymorphisms) defining different branches of the Y-chromosome tree have been discovered from these comparisions of 23andMe Y-DNA data.

    There are a number of additional tools which you can use to explore your raw data. You will find a full list here:


  5. Great review of 23andme. I was sitting on the fence about buying a kit myself on DNA day when the sale was on. Now I really wish I did go for it. Oh well, I am sure I will get another chance.

    Since there are gene patents on some SNPs did you find any extra details on SNPedia?

  6. I’m intruiged.
    But a bit uneasy about letting megacorporation handle my data.
    How do SNPedia/Promethease analyses/results compare with 23andme?


  7. Morgan, Yannick – yes, Promethease/SNPedia returns a lot of extra detail. I ran it last night, it annotated 6730 genotypes, so I’m still wading through the report. There seems to be good overlap with the 23andMe reports.

  8. Pingback: Health 2.0 News: JMIR, Wi-Fi and IF « ScienceRoll

  9. 23andme has great value, but the fact that you don’t own your data is too intrusive.
    And after it was bought by google it’s just crystal clear.
    Thanks, but no thanks.

    • Google have invested in 23andMe, I don’t believe that they own it. Data ownership and privacy are concerns for some people – and those people can choose not to participate.

      • It’d be very scary if people wasn’t allowed to choose not to participate :).
        Data ownership and privacy regarding your own DNA seems pretty important, not only for paranoids. Insurance companies would be interested on anyone’s health data.

        For me, it’s like a trojan horse: you get fancy charts and the false illusion that your data is of no interest to anyone but you, or is private, or is no one else’s business, but meanwhile they collect as much data as possible. Compare health data with web traffic, which has received lots of effort spent on collecting it. It’s like not thinking twice before using as your new DNS server. Sure, it’s free, and easy to remember.

        It’s not a matter of being paranoid or not. It’s a matter of ignoring the value underneath of lots of people like yourself careless ignoring where your data is, who could be using it, and what for. Maybe it doesn’t matter now, but it’ll matter soon.

        • I have a feeling you don’t have a Facebook account? ;)

          Of course these concerns are well known, but the privacy statement on the 23andme website is very clear:
          “23andMe respects your privacy. 23andMe does not sell, lease, or rent your individual-level Personal Information without explicit consent.” https://www.23andme.com/about/privacy/

          Also, (in the US at least) the Genetic Information Non-Discrimination Act prevents health insurance or employers from using genetic information to discriminate anyone. https://www.23andme.com/you/faqwin/gina/

          I haven’t taken the test yet, but privacy concerns are very low on the reasons why I haven’t yet.

Comments are closed.