Thursday, 18 June 2015

23andMe genotypes one millionth customer

23andMe have today announced that they have genotyped their one millionth customer. The full press release can be read here. 23andMe have also sent out personal e-mails to all their customers advising them of their customer number. I am customer number 46,957. I was one of the early adopters in the UK. I couldn't afford to pay the then price of $499 but I took the opportunity to order a test for $99 when there was a flash sale on DNA Day in April 2010. There is a certain kudos to having a five-digit customer number but we do have a number of people in our genetic genealogy community in America who have four-digit numbers, and even some genetic genealogy pioneers with coveted three-digit numbers! I've copied below the e-mail that I received from 23andMe. The content of the e-mail is also reproduced on the 23andMe blog.

Subject: Reaching 1 million. You're #46,957
If this email isn’t displaying correctly, view it in your browser.

Last week, we genotyped our one millionth customer. You are part of the one million people driving change.

One million is more than a number. It's a turning point. We are taking control of our data. We are taking ownership of information about ourselves. We believe knowing more about who we are can benefit society, not just the individual.

Just fifty years ago, doctors were reluctant to tell patients if they had cancer.1 The world is different today.

One million customers ago, attendees were shocked at the annual American Society of Human Genetics conference when direct-to-consumer genetic testing services were announced.2

One million customers ago, we didn't know how consumers would react to direct-to-consumer testing, but now there are studies showing consumers don't overreact to their results.3

One million customers ago, we didn't have the Genetic Information Non-Discrimination Act in the United States or other personal data protection laws that are now in place in many countries around the globe.

One million customers ago, we didn't have direct access to our health laboratory results. The United States recently mandated individuals can get laboratory test results directly from the laboratory upon request.4

As customer number 46,957, you are part of this unique group of one million people driving change. I celebrate you, your 23andMe story and the power of all of us today: #PowerOf1Million.

You'll often overhear me enthusiastically telling people to "Spit!" but today is the day to be proud and let everyone know: "I Spat!"


Anne Wojcicki
CEO and Customer #60
You are 23andMe
customer number


You are receiving this email because you are a customer of 23andMe.

Please add to your address book to ensure our emails reach your inbox.

23andMe, Inc. uses the clinical laboratory services of
National Genetics Institute, a subsidiary of Laboratory Corporation,
2440 S Sepulveda Blvd., Ste #130, Los Angeles, CA 90064.

© 2007-2015 23andMe, Inc. 899 W. Evelyn Avenue. Mountain View, CA 94041
update your 23andMe email preferences | unsubscribe from 23andMe news and updates emails

23andMe are now the first DNA testing company to reach the milestone of having one million people in a single database.

AncestryDNA are probably also very close to the one million mark. They currently state that they have 850,000 people in their database, but they have been citing this figure for some time. They have previously said that they are selling 150,000 kits per quarter. We can probably expect an announcement very soon.

Family Tree DNA were the first company to hit the one million milestone but their tally of one million test takers is spread across two different databases. FTDNA have their own proprietary database but they also do the testing for the Genographic Project. There is some overlap between the two databases as Genographic customers can transfer their results to the FTDNA genealogical matching database.

It's taken nearly eight years for 23andMe to reach the one million milestone but the big growth only started to take place when they dropped the price of their test to $99 in December 2012. I wonder how long we will have to wait before we see the first two million or five million consumer DNA database.

Related blog posts
- What is the current size of the consumer genomics market?
- 23andMe relaunches health reports in the UK
- My series of articles on my 23andMe test
- Autosomal DNA testing now affordable for all

Wednesday, 10 June 2015

The AncestryDNA international roll-out, shaky leaves and AncestryHealth

The AncestryDNA test is now starting to be rolled out in other countries. The test was officially launched in the US in May 2012. It became available in the UK and Ireland in January 2015, and was launched in Australia and New Zealand at the end of May. The test became available in Canada this week. It is not known if Ancestry will sell the test in any other countries. They also have offices and a subscriber base in Sweden and Germany. Germany has restrictive laws on direct-to-consumer testing so it seems unlikely that the test will be offered there. However, Sweden might well be another target market for the company in the near future.

It is interesting to note that the AncestryDNA test is much more expensive outside the US, and it remains to be seen how this will impact the take up of the test in other countries. Here is a comparison of the prices.

US $99 + $9.95 for shipping = US $108.95
Australia and New Zealand AUS $149 + $29.99 for shipping = $178.99 or US $138
Canada CAN $149+ $19.95 for shipping = $168.95 or US $138.40
UK and Ireland £99 + £20 shipping = £119 or US $183

Note that we are paying much more for the test in the UK and Ireland than in any other country. However, Ancestry do have frequent promotions such as the current offer of 10% off for Father's Day so do look out for these special offers.

I put myself on the invitation list when the AncestryDNA test first came out in the US, and was able to order a test in June 2012 despite the fact that I don't live in America, though this loophole was quickly closed. The AncestryDNA database has grown at a phenomenal rate and by the end of March this year they had DNA results for over 850,000 people. However, with a database that has until this year been 99.9% American I've had very little out of the test so far. I have no known cousins in America. Nevertheless I have had a lot of matches but they are nearly all with very distant cousins with deep roots in Colonial America. They generally have no surnames or geographical locations in common with me and it would be an impossible task trying to find the genealogical connection in these circumstances.

Now that the first results are starting to come in from the UK the situation is changing and I'm pleased to report that I have my first ever shaky leafy hint at AncestryDNA which has enabled me to confirm a genealogical connection with a third cousin in the UK. The shaky leaf hints are well named because the leaf really does shake when you first see it! This is what my match page now looks like.

The predicted relationship for my new match was third to fourth cousin. When I click to view the match I get taken to a page where I can see the shared ancestor hint and an outline of our two trees showing our connection to the common ancestral couple, my great-great-grandparents Charles James Wiggins and Mary Ann Thorn. (I've blurred out the names of the recent generations in the screenshot below.)

In this particular case the shaky leaf hint has worked exceptionally well, and my match turned out to be a third cousin. Because it's a close relationship it probably would have been a straightforward matter finding the connection but it certainly helps having Ancestry doing all the hard work for you. I shall look forward to confirming more relationships as more results start to come back from the UK and Ireland. I also look forward to receiving matches with people in Australia, New Zealand and Canada, as I have a number of links with all three countries in my family tree.

The frustrating part of the AncestryDNA test is that there is no way of verifying the validity of the match. Ancestry do not provide the underlying segment data on which the match is based and they don't offer a chromosome browser. It is rather like being given access to an index and then finding that the key records that you want in the index have been withheld for no good reason and you are expected to rely instead on someone else's transcription. I am hoping to persuade my match to transfer her results to Family Tree DNA using the autosomal DNA transfer programme. I'm also hoping that she will be able to transfer her results to GedMatch. However, we should not have to rely on transferring our data to other websites in order to access such basic data. I hope that Ancestry will fix this issue as a matter of priority.

In another development Roberta Estes reports that Ancestry are now beta-testing a new AncestryHealth website in the US. Roberta has bravely volunteered as a guinea pig and tried out the new service. I suggest you read her blog post if you want to find out more. It remains to be seen if AncestryHealth will become available in other countries. It is not clear to me what benefits can be derived from such a service as it is not telling you anything that you didn't already know. There are already a number of free tools available online to calculate your risk of developing various conditions. Here are three examples that I found in a quick internet search: Stroke Risk Calculator, Diabetes Risk Calculator, and Oncolink. However, if I had a family history of a particular condition I would much rather go and visit my GP to determine my risk rather than input the information on a website. If I want to record my own health information online I can use a free service like the Microsoft HealthVault.

Although AncestryHealth is currently a separate entity from AncestryDNA you can import your family tree and also your ethnicity estimates from AncestryDNA. I noticed that the AncestryDNA Privacy Statement and Terms and Conditions were updated on 18th May and the changes take effect for existing customers from 18th June 2015. I have not received an e-mail from Ancestry advising me of these changes but the updates were visible when I checked into my AncestryDNA account. The Privacy Statement was also updated back in February 2015 though I don't recall noticing this update at the time. On comparing the different versions the biggest change I noticed was introduced back in February this year. AncestryDNA now make it explicitly clear that users' DNA results will be internally analysed not just for genealogical purposes but also for "medicine and other topics".

Here is the relevant clause from the old Privacy Statement dated 20th March 2013:
v) To Research the human genome. If you voluntarily agreed to the AncestryDNA Consent Agreement we may use the Results and other information in an aggregated and anonymous form, for the purposes of research and publication and in accordance with the AncestryDNA Consent Agreement.
Ancestry do state here that your DNA results can be used for research but the implication is that this will only happen if you voluntarily sign the AncestryDNA Consent Agreement. Contrast this clause with the updated Privacy Statement dated 20th February 2015:
vi) To perform research: AncestryDNA will internally analyze Users’ results to make discoveries in the study of genealogy, anthropology, evolution, languages, cultures, medicine, and other topics. In addition, if you voluntarily agreed to the Research Project Informed Consent we may use the Results and other information for the purposes of collaborative research and publication and in accordance with the Informed Consent.
If you sign up for AncestryHealth you are given the option to sign an Informed Consent form to participate in research. Similarly, if you take an AncestryDNA test you are also given the option to participate in research by signing an Informed Consent form. The wording of both forms is almost identical. Here is the description of the Research Project from AncestryDNA:
1. What is the research project?
The Ancestry Human Diversity Project collects, preserves and analyzes genealogical pedigrees, historical records, surveys, family health data, medical and health records, genetic information, and other information (collectively, "Information") from people all around the world in order to conduct research studies to better understand, among other things, human evolution and migration, population genetics, population health issues, ethnographic diversity and boundaries, genealogy, and the history of our species ("The Project"). Researchers hope that the Project will be an invaluable tool for a wide range of scholars interested in genealogy, anthropology, evolution, languages, cultures, medicine, and other topics and that the Project may benefit future generations.
Here is the description of the project from AncestryHealth:
What is the research project?
The Ancestry Human Diversity Project utilizes genealogical pedigrees, historical records, surveys, family health data, medical and health records, genetic information, and other information (collectively, “Information”) from people all around the world in order to conduct research studies to better understand, among other things, human evolution and migration, population genetics, population health issues, ethnographic diversity and boundaries, genealogy, and the history of our species (“The Project”). Researchers hope that the Project will be an invaluable tool for a wide range of scholars interested in genealogy, anthropology, evolution, languages, cultures, medicine, and other topics, and that the Project may benefit future generations.
Many people will have signed the Informed Consent form on AncestryDNA without realising what they were doing as it is presented as just one extra box to be ticked to complete the order process, and it is easy to tick the box without even reading the form. I wrote of my concerns about this process when I activated my own kit back in 2012. At that time the research project was known as the Human Genome Diversity Project. As I was unable to find out any further information about what this project involved I did not sign the form and have still not done so.

I suspect many people who took the AncestryDNA test will be surprised to find that by submitting their DNA for a genealogical DNA test their results might also be used for medical research. I'm quite happy to have my DNA used for research purposes. I've tested at 23andMe, and I am participating in the UK Biobank Project which is sequencing the DNA of over 500,000 British people. However, if people's DNA results are to be used for medical research I do think that the process should be transparent and that informed consent should be obtained up front rather than changing the terms and conditions and the privacy policy and hoping that people won't notice.

Despite these concerns I do think it is worthwhile doing the AncestryDNA test. Ancestry have a big marketing machine and will inevitably attract customers who would not have considered taking a test anywhere else. However, if you are taking a test to confirm a specific hypothesis about relationships then the Family Finder test from Family Tree DNA is the most cost-effective option. The Family Finder test costs around £70 inclusive of postage versus £119 for the AncestryDNA test.

For a detailed comparison of all the different autosomal DNA testing companies see Tim Janzen's Autosomal DNA testing comparison chart in the ISOGG Wiki. See also Sue Griffith's Notes for UK (and ex-US) residents on DNA testing companies where she provides advice and detailed costings on testing with all the different companies from a UK perspective.

© 2015 Debbie Kennett

Thursday, 21 May 2015

The business of genetic ancestry

There is a programme going out next Tuesday 26th May on BBC Radio 4 on the "business of genetic ancestry". The programme is presented by Dr Adam Rutherford and includes interviews with Professor Mark Jobling from the University of Leicester, my colleague Professor Mark Thomas from University College London and yours truly! You can find further details about the programme here:

The programme will also be available on the iPlayer and it will be repeated on Monday 1st June at 9pm.

I paid a visit to New Broadcasting House at the end of last month to record my interview with Adam. I was interviewed for one hour and 20 minutes, and we had a wide-ranging conversation about the exciting discoveries that can be made from DNA testing but also about some of the problematic press coverage and the dubious claims made by certain companies. Obviously only a tiny fraction of what I said will make the final cut, so I shall be interested to see how it turns out.
In the recording studio with Adam Rutherford.
I have been very impressed with the way that the programme has been made. The BBC have done a lot of research behind the scenes and have gone to great pains to talk to a wide range of people and to present a balanced view showing both the benefits of genetic ancestry testing but also highlighting some of the problematic areas. They invited people from all sides of the debate to contribute to the programme though I understand that unfortunately not everyone who was invited to participate chose to do so.

If you listen to the programme do let me know what you think.

Related blog posts
- Driving in the wrong direction with a dodgy DNA satnav
- More on the S4C DNA Cymru controversy and my review of "Who are the Welsh?"

© 2015 Debbie Kennett

Sunday, 17 May 2015

Three generations of FTDNA MyOrigins admixture results

I wrote last week about comparing my admixture results from Ancestry, 23andMe and Family Tree DNA. At Family Tree DNA I have now tested three generations of my family so I thought it would be interesting to compare the MyOrigins results across the generations. I've provided below a summary of the ancestry for each person tested together with a screenshot of their results. Click on the images to enlarge them.

Debbie's dad
Four grandparents born in England: Bristol, Gloucestershire, London (x2).

Eight great-grandparents born in England: Bristol (x2), Devon, Essex, Gloucestershire, Hertfordshire, London (x2).

Fifteen great-great grandparents born in England: Devon (x2), Bristol, Essex, Gloucestershire, Hertfordshire (x 2), London.
One great-great grandparent born in Scotland (location not known).
The birthplace of seven of his English great-great-grandparents is unknown. Four were probably born in Bristol or in a nearby county. Three were Londoners who could have moved to London from anywhere in England.

Debbie's mum
Four grandparents born in England: London (x2), Hampshire (x2).

Eight great-grandparents born in England: Berkshire, Hampshire, London (x3), Somerset, Wiltshire. The birthplace of one great-grandparent is not known but he was probably born in London.

Fifteen great-great-grandparents born in England: Bedfordshire, Berkshire (x2), Gloucestershire, Hampshire (x2), Hertfordshire, London (x2), Somerset (x2), Wiltshire.
One great-great-grandparent born in Ireland: County Kerry.
The birthplace of three of her English great-great-grandparents is unknown. One was probably born in Hampshire. The other two were probably Londoners who could have come from anywhere in the country.

Four grandparents born in England: Bristol, London (x3).

Eight great-grandparents born in England: Bristol, Gloucestershire, Hampshire (x2), London (4),

Sixteen great-great-grandparents born in England: Berkshire, Bristol (2), Devon, Essex, Gloucestershire, Hampshire, Hertfordshire, London (x 5), Somerset and Wiltshire.
The one great-great-grandparent with an unknown birth location was probably born in London.

Twenty-four great-great-great grandparents born in England: Berkshire (x2), Bristol, Devon (x2), Essex, Gloucestershire (x2), Hampshire (x2), Hertfordshire (x3), London (x5), Somerset (x2), Wiltshire.
One great-great-great grandparent born in Ireland: County Kerry
One great-great-great grandparent born in Scotland (location not known).
The birthplace of the remaining eight English great-great-great-grandparents is unknown but they were probably born in Bristol, London and Hampshire.

Debbie's husband
Four grandparents born in England: Cambridgeshire (x2), Cumberland, Devon.

Eight great-grandparents born in England: Cambridgeshire (x3), Devon (x2), Dorset, Somerset, Surrey.

Sixteen great-great grandparents born in England: Cambridgeshire (x3), Devon (x4), Hampshire, Herefordshire, Hertfordshire, Huntingdonshire (x2), Somerset (x2), Surrey (x2). 

Twenty-six great-great-great grandparents born in England: Cambridgeshire (6), Devon (x8), Hampshire, Herefordshire (x2), Huntingdonshire, Somerset (x4), Surrey (x3), Sussex.
The birthplace of his remaining six  English great-great-great-grandparents is unknown. Three were probably born in Cambridgeshire, two in Hertfordshire and one in Surrey.

Debbie's eldest son

As can be seen, there is considerable variation between family members. This is only to be expected because of the random nature of DNA inheritance. However, some of the differences are somewhat more extreme than might be intuitively expected. For example, 57% of my DNA matches the British Isles cluster whereas only 40% of my dad's DNA matches the British Isles and only 7% of my mum's DNA.

My dad, my husband and my son all come out with smaller percentages of "Middle Eastern" DNA. My husband and son's "Middle Eastern" DNA appears on the map over Turkey, Georgia and Azerbaijan whereas my dad's supposedly Middle Eastern DNA is centred over Egypt and Jordan. I've noticed that a significant proportion of the members of my Devon DNA Project with predominantly British ancestry are coming out with these small percentages of "Middle Eastern". Clearly this does not mean that any of them have recent ancestry from the Middle East, and it is probably related to the limitations of the available reference populations. I hope to look at the Middle Eastern issue in more depth in a subsequent blog post.

As I mentioned in my previous blog post, these admixture results should really only be used for entertainment value at present. However, the results are likely to change over time as more reference populations become available and as the methodology improves. Admixture tests should really be regarded as a bonus feature of an autosomal DNA test and should not be the primary purpose for testing. While your admixture results will not tell you whether your great-great-grandfather was Scottish or Irish, you might find instead that you match with a cousin who is descended from the ancestor of interest who will be able to fill in the blanks in your family tree. Cousin-matching tests will become increasingly useful as the databases continue to grow in size.

Related blog posts

© 2015 Debbie Kennett