23andme – yes, me – part 3

Results!
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.

Overview

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
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.