More people bought commercial DNA tests in 2017 than in all previous years combined

[Maciamo]

Technology Review: 2017 was the year consumer DNA testing blew up

"The number of people who have had their DNA analyzed with direct-to-consumer genetic genealogy tests more than doubled during 2017 and now exceeds 12 million, according to industry estimates.

Most of those tested are in the US, suggesting that around 1 in 25 American adults now have access to personal genetic data—a figure that could spur a range of new genetic analysis services.

The boom comes amid a price war in which companies offered under-$60 tests and 2-for-1 deals during an end-of-year blitz of advertising and discounts."


The article mentions that AncestryDNA is now the biggest DNA testing company, with 7 million customers, well ahead of 23andme with a bit over 3 million.

That's rather odd since AncestryDNA doesn't test either Y-DNA or mtDNA and does not offer the very interesting Health Report provided by 23andMe (to American customers at least). It's richly ironic than a test designed for genealogists shouldn't test Y-DNA, which is the main (and possibly the only useful) tool of genetic genealogists to determine the number of generations separating two individuals descended from a common paternal ancestor.

AncestryDNA also has rather mediocre autosomal report (unless they have changed it recently without my knowing it), which made me rank it the worst DNA testing company in my review of ancestral DNA tests last year. Only FTDNA's Family Finder scored worse, but FTDNA does offer other tests to make up for it.

It looks like the majority of Americans prefer a slightly cheaper but mediocre product rather than go for quality. The same is true of food, TV, English grammar, and more other things. If so many people are satisfied with mediocrity it is no wonder than half of the US population is now obese and close to or under the poverty line. Really sad. I suppose that in this context the success of AncestryDNA is not that surprising.

[Jovialis]

1 in 25 is an impressive number, and it will probably increase. Also, it is indeed odd and unfortunate that AncestryDNA wouldn't test for Y-DNA and mtDNA, considering the fact it sold the most. 23andme is clearly the better company, with more to offer.

I bet these companies experienced a surge in sales right before the holidays, since DNA testing kits were a popular Christmas gift. I myself had purchased 23andme kits for my whole family.

[elghund]

Technology Review: 2017 was the year consumer DNA testing blew up

"The number of people who have had their DNA analyzed with direct-to-consumer genetic genealogy tests more than doubled during 2017 and now exceeds 12 million, according to industry estimates.

Most of those tested are in the US, suggesting that around 1 in 25 American adults now have access to personal genetic data—a figure that could spur a range of new genetic analysis services.

The boom comes amid a price war in which companies offered under-$60 tests and 2-for-1 deals during an end-of-year blitz of advertising and discounts."


The article mentions that AncestryDNA is now the biggest DNA testing company, with 7 million customers, well ahead of 23andme with a bit over 3 million.

That's rather odd since AncestryDNA doesn't test either Y-DNA or mtDNA and does not offer the very interesting Health Report provided by 23andMe (to American customers at least). It's richly ironic than a test designed for genealogists shouldn't test Y-DNA, which is the main (and possibly the only useful) tool of genetic genealogists to determine the number of generations separating two individuals descended from a common paternal ancestor.

AncestryDNA also has rather mediocre autosomal report (unless they have changed it recently without my knowing it), which made me rank it the worst DNA testing company in my review of ancestral DNA tests last year. Only FTDNA's Family Finder scored worse, but FTDNA does offer other tests to make up for it.

It looks like the majority of Americans prefer a slightly cheaper but mediocre product rather than go for quality. The same is true of food, TV, English grammar, and more other things. If so many people are satisfied with mediocrity it is no wonder than half of the US population is now obese and close to or under the poverty line. Really sad. I suppose that in this context the success of AncestryDNA is not that surprising.

23andMe's ydna and mtdna haplogroup info can disprove but not prove paternal-line relationships. Only FTDNA, with its y-dna STR tests, can do that for customers right now. AncestryDNA's test's biggest advantage is its enormous database of tested customers. Plus, AncestryDNA's customer base is the most interested in genealogy. Their customers tend to share family trees and explore common ancestry through "DNA Circles."

As far as 23andMe's Health Info, they don't offer much, since the FDA stifled that aspect of their service. Right now, they provide Wellness info for eight areas:Genetic Weight; Lactose Intolerance; Alcohol Flush Reaction; Caffeine Consumption; Deep Sleep; Muscle Composition; and Sleep Movement. The test results given do not give definitive answers but give the More like to" or "less likely to." A survey of the forums of where people actually discuss the accuracy of their responses reveals accuracy about half the time, what you might expect with randomness. As far as the Traits, 22 are listed, most of which you can determine with self-examination with no test at all.

The genealogists are keeping all these companies alive and most of them are stifling their communication in an assortment of ways. Those who order any of these test for health info will be sorely disappointed. My guess, over half of these genetic companies will be history within ten years, and old customers will probably not even be able to log into them anymore. Ancestry DNA is running a huge lead in numbers and in genealogical interest right now. FTDNA has their niche for Y-DNA STR testing. Plus, it was smart move of their behalf to allow free uploads. Living DNA is getting the corner on the European market. MyHeritage is getting off to a good start, but it's like a sapling under a dense canopy of trees. It seems to me, 23andMe's is heading the direction of Big Pharma becoming its largest customer, and offers less and less to the genealogist, the backbone of the genomics industry. They've taken away from their service in the past few years compared to what they have added. I think all these sales, which started before Christmas and extended again and again through much of February, show a sudden desperation. When FTDNA and 23andMe dropped their prices $20, Ancestry DNA dropped theirs $30. Right now, Ancestry DNA has the biggest database and they are growing the fastest. Most customers for these tests want either one or both of two- things-- ethnicity reports and to find cousins and ancestors.

Ancestry DNA right now is providing a value, though I wouldn't put it past them one day to lock everything up behind paid subscription services, like they gradually did with their free family trees. I think a person's best bet, if they have the expendable income and interest in genetic genealogy, is to purchase tests from multiple companies. Different sources are going to give you different perspectives. Ancestry DNA currently offers the greatest number of second and third cousin matches and beyond and the best response rates from customers seeking to compare information. They are not without a competitive edge.

[Maciamo]

23andMe's ydna and mtdna haplogroup info can disprove but not prove paternal-line relationships. Only FTDNA, with its y-dna STR tests, can do that for customers right now.

That's not true. The best test to prove a paternal-line relationship would be Full Genome Corp's Y Elite Ancestry test, which covers most of the Y chromosome (much more than FTDNA's BigY, and no comparison to a lame STR test). Another alternative is to order SNP panels either from FTDNA or the cheaper YSEQ.

AncestryDNA's test's biggest advantage is its enormous database of tested customers. Plus, AncestryDNA's customer base is the most interested in genealogy. Their customers tend to share family trees and explore common ancestry through "DNA Circles."

I don't see why the big database is an advantage. It only helps to find potential distant cousins, but my experience with 23andMe showed that these people are not possibly related in genealogical times (not 5th cousins or less as they claim). So what's the point? I think that distant relative search is one of the biggest scams of the DNA testing industry.

As far as 23andMe's Health Info, they don't offer much, since the FDA stifled that aspect of their service. Right now, they provide Wellness info for eight areas:Genetic Weight; Lactose Intolerance; Alcohol Flush Reaction; Caffeine Consumption; Deep Sleep; Muscle Composition; and Sleep Movement. The test results given do not give definitive answers but give the More like to" or "less likely to." A survey of the forums of where people actually discuss the accuracy of their responses reveals accuracy about half the time, what you might expect with randomness. As far as the Traits, 22 are listed, most of which you can determine with self-examination with no test at all.

23andMe tests over 50 carrier status for genetic diseases, and the answers are definitive. Likewise some reported alleles for Alzheimer's or some cancer risks (BRCA gene for women) do confer a very high risk to the carriers, so it's useful to know. Traits are just for fun as those are things you already know about yourself.

The genealogists are keeping all these companies alive and most of them are stifling their communication in an assortment of ways.

But why? Autosomal tests are pretty useless for genealogy unless someone was adopted or doesn't know anything about their close relatives (maybe one person out of 100,000). Tests like AncestryDNA, Family Finder, MyHeriage and 23andMe are all useless for genealogy. You can't trace back ancestral lines without a detailed Y-DNA test (minimum FTDNA's 111 STR, or better FGC's Y Elite test).

It's baffling why millions of people order those tests. Personally I don't care much about DNA for genealogy anyway. I think there are two other very good reasons to order DNA tests:

1) To know about one's distant ancestry (e.g. what admixtures we inherited from ancient ethnic groups, what's our haplogroups)

2) Knowing about one's health.

Unfortunately no testing company offers a look into ancient autosomal ancestry at present. 23andMe and LivingDNA have some admixtures that can be used as proxies for some ancestral populations (Germanic/Scandinavian, Italian/Roman, Basque, etc.), but they aren't very accurate yet.

23andMe offers a pretty good deal for health reports considering that more can be obtained through SNPedia or other websites. Many other DNA companies charges many times more (often over $1000) for tests that may have more detailed reports but do not test more useful SNPs. So 23andMe remains one of the best choices IMO.

[Dibran]

..........................................

I wouldnt say autosomal tests are useless, but rather not the most important. Just an integral part of the whole in the genetic picture so to speak. I do agree Y-DNA is important, but it makes up minimal portion of the genome. For instance with Y-Elite I was confirmed R1a-M458-L1029* negative downstream and 24 unigue SNPs. I was suspected to form a founder effect, a "Albanian" sub-branch under L1029. My only "close" match ordered a SNP panel for M458. He is also Albanian(from Gostivar, Macedonia) with a TMRCA between us of 1000-1600 years. The line likely implanted itself and due to the isolation caused the founder effect?

But, so far its only the two of us forming our own cluster. I feel the reason basic autosomal tests are the most sought after, is due to the cheap costs. The average person whose interested in genetics, but not enough to contribute towards unraveling deeper, are less likely to pay 500-700 dollars for a BigY or Yelite.

The costs need to go down to increase the testing for Maternal/Paternal Haplogroups. I keep hearing it may eventually become as cheap as 100 dollars. If that is the case, I would be ordered more than a few tests to test further members of my family and close friends.

The problem is, no one wants to spend an arm and a leg to learn this stuff. I had to set aside money little by little to afford my Yelite.

[mwauthy]

I think these numbers also show the power of advertising. Ancestry and 23andMe are the only companies that advertise regularly on American tv, and Ancestry by far runs more commercials than 23andMe at least according to what I see.
I think a few years ago I might have seen the occasional Natgeo commercial and I think sometimes I see the occasional Myheritage advertisement. I don’t think I’ve ever seen an FTDNA commercial. So for the average American who puts no research into their options Ancestry and 23andMe are their only apparent choices.
I also think these companies have fooled many Americans who are heavily European admixed into believing that these tests will accurately specify their ancestors’ exact countries of origin (trading lederhosen for a kilt). This is appealing to many people who have centuries of American Ancestry with an incomplete family tree.
I don’t think the average American cares if they are related to “John Doe” 500 years ago. I think they want to know how much Italian or English they are even though with the current technology this is not possible. The companies have created an illusion that this is possible leading many people to waste a lot of time searching for phantom ancestors that match their ethnicity percentages.
Real genealogical research through detailed y-Chromosome analysis to find snps in the time frame since surname creation is still too pricy for most people averaging around $500. If future companies can drop the price to somewhere between $100-$200 I think y-chromosome analysis will really take off and it will be very interesting to pinpoint when lineages in various countries split away from each other.

[Wheal]

I think the ethnicity tests are questionable. Here is a sample from Family Tree results run through gedmatch using the k13 calculator

Attachment 9765

[mwauthy]

I think the ethnicity tests are questionable. Here is a sample from Family Tree results run through gedmatch using the k13 calculator

Attachment 9765

They are very questionable. The best they can do for Europeans is separate Northern European alleles from Southern European alleles. Your Northern European alleles can be assigned anywhere from Ireland to Finland and your Southern European alleles can be assigned anywhere from the Iberian Peninsula to the Caucasus.

[Maciamo]

I think these numbers also show the power of advertising. Ancestry and 23andMe are the only companies that advertise regularly on American tv, and Ancestry by far runs more commercials than 23andMe at least according to what I see.
I think a few years ago I might have seen the occasional Natgeo commercial and I think sometimes I see the occasional Myheritage advertisement. I don’t think I’ve ever seen an FTDNA commercial. So for the average American who puts no research into their options Ancestry and 23andMe are their only apparent choices.
I also think these companies have fooled many Americans who are heavily European admixed into believing that these tests will accurately specify their ancestors’ exact countries of origin (trading lederhosen for a kilt). This is appealing to many people who have centuries of American Ancestry with an incomplete family tree.
I don’t think the average American cares if they are related to “John Doe” 500 years ago. I think they want to know how much Italian or English they are even though with the current technology this is not possible. The companies have created an illusion that this is possible leading many people to waste a lot of time searching for phantom ancestors that match their ethnicity percentages.
Real genealogical research through detailed y-Chromosome analysis to find snps in the time frame since surname creation is still too pricy for most people averaging around $500. If future companies can drop the price to somewhere between $100-$200 I think y-chromosome analysis will really take off and it will be very interesting to pinpoint when lineages in various countries split away from each other.

I completely agree with this.

They are very questionable. The best they can do for Europeans is separate Northern European alleles from Southern European alleles. Your Northern European alleles can be assigned anywhere from Ireland to Finland and your Southern European alleles can be assigned anywhere from the Iberian Peninsula to the Caucasus.

However I disagree here. It is possible now to distinguish more than Northern European vs Southern European admixtures. Irish and Finnish admixture are very distinct from each others. Living DNA can even tell apart (although not with 100% confidence) ancestry for regions within Britain, and soon that should be extended to all regions of Europe, and later to other parts of the world. Things are changing fast.

[mwauthy]

I completely agree with this.



However I disagree here. It is possible now to distinguish more than Northern European vs Southern European admixtures. Irish and Finnish admixture are very distinct from each others. Living DNA can even tell apart (although not with 100% confidence) ancestry for regions within Britain, and soon that should be extended to all regions of Europe, and later to other parts of the world. Things are changing fast.

I agree that some companies have finer tuned algorithms and more complete reference samples than other companies, and I have a lot of hope in Living DNA’s future regional breakdowns in other places besides the UK.

Speaking from experience though after taking over 10 tests and running my parent’s and my wife’s data through various tests, I’ve seen huge discrepancies in certain regions across Europe. The only way to explain this is that many alleles are found all over Europe and which region it gets assigned to can be arbitrary based on the reference samples and algorithms.

Here are some examples:

French Canadian Mother:

Ancestry: <1% Europe West
FTDNA: 99% West and Central Europe

FTDNA: 0% British Isles
MyHeritage: 13.6% Irish, Scottish, And Welsh


Wallonia Belgian Father:

FTDNA: 0% English
MyHeritage: 53.2% English

FTDNA: 1% Finnish
DNA Land: 7.8% Northeast European
Geneplaza K29: 13% Baltic


Half Italian Wife:

23andMe: 32.5% Italian
FTDNA: 32% Iberia

[AdeoF]

It's also possibility due to people saying how great it is to look at there DNA and the youtube videos that shows that aswell which ups the advertisements. Also a lot of people are interested what part of the world there DNA matches with.

[Salento]

MIT: More than 26 million people shared their DNA with ancestry firms, allowing researchers to trace relationships between virtually all Americans:


...More than 26 million people — more people than all of Australia — have shared their DNA with one of the four leading ancestry and health databases, allowing researchers to extrapolate data on virtually all Americans...

https://www.cnbc.com/2019/02/12/priv...try-firms.html

[gidai]

MIT: More than 26 million people shared their DNA with ancestry firms, allowing researchers to trace relationships between virtually all Americans:


...More than 26 million people — more people than all of Australia — have shared their DNA with one of the four leading ancestry and health databases, allowing researchers to extrapolate data on virtually all Americans...

https://www.cnbc.com/2019/02/12/priv...try-firms.html

26 milion people! For entire World, that would mean an average of 140,000 / country. Mostly big countries should have a much larger number. The available studies refer to a number hundred times smaller, mostly of only several tens or hundreds of individuals! There should now be studies done on hundreds of times more samples, which would greatly enhance the accuracy of statistics results.
Where are available studies made on such a large number of people who have given their consent !?

[Farstar]

I have asked the same question in another post, but I think it makes sense to ask the same here:

Is there a way to do a DNA test at home, without having to send the data to any company? If this is not possible right now, is there a estimation of when it will be possible?

Of course, a related question is how the data could be interpreted and analyzed (even for health purposes), if one could do a DNA test at home.

Note: I want a DNA testing in order to understand my "old origins", and in order to understand potential health issues. I am not interested (so much) about genealogy (at least, if this implies to share data with others).

[Salento]

26 milion people! For entire World, that would mean an average of 140,000 / country. Mostly big countries should have a much larger number. The available studies refer to a number hundred times smaller, mostly of only several tens or hundreds of individuals! There should now be studies done on hundreds of times more samples, which would greatly enhance the accuracy of statistics results.
Where are available studies made on such a large number of people who have given their consent !?

26 million is more than the population of Australia.

The article also says that MIT predicts that in 2 years that number will rise to 100 million.