Calling HOGWASH on 23andMe’s Ancestry Timeline

Every now and then, I’m aghast when I look at a product and wonder how the devil it ever escaped the lab.  Is there no quality control?  And who thought it was a good idea, anyway, and why?

23andMe’s new Ancestry Timeline, released last week, is one of those.

Not only is it incorrect, but it deceives people into believing something that isn’t true.

Let’s take a look.

23andme-timeline

My Ancestry Timeline at 23andMe is shown above. I notice that my Middle Eastern/North African is missing from the timeline.  It’s less than 1%, but then so is my Native American which is included.

You can see in the text underneath the timeline that 23andMe says this timeline reflects how long ago my MOST RECENT ancestor in that geographic location was born.

Let’s compare this with reality.  You may recall that I recently wrote the article, Concepts – Calculating Ethnicity Percentages. In that article, I utilized my known and proven genealogy for my 64 great-great-great-great-grandparents to calculate what my ethnicity results should look like.  I’m referring to the same chart of my 64 ancestors for this exercise as well, since I’ve already done a great deal of the work.  Let’s see how reality stacks up to the 23andMe timeline.

On the chart below, I’ve shown the geographic category, the dates from the 23andMe timeline reflecting my most recent ancestor’s birth, my most recent ancestor from that location, and the accuracy of the 23andMe estimate.

Category 23andMe Dates My Most Recent Ancestor Birth 23andMe Accuracy
British and Irish 1900-1930 1759 – Henry Bolton Utter hogwash
French and German 1840-1900 1854 – Hiram Ferverda Close
Scandinavian 1750-1840 No ancestor More hogwash
Eastern European 1720-1810 No ancestor Hogwash
Italian 1690-1810 No ancestor Hogwash
Native American 1690-1790 Uncertain, mother’s side – early 1600s, father’s side – unknown Not verifiable, reasonable

The part of this equation that I find extremely upsetting is the sheer magnitude of how misleading the 23andMe timeline is.  It’s not just wrong, it’s horribly deceptive – massively inaccurate by any measure possible.

Here’s what the 23andMe white paper says about this new tool:

“Admixture date estimator is a 23andMe feature that enables customers to find out, for each of the ancestries they carry, when they may have had an ancestor in their genealogy who was likely to be a non-admixed representative of that population.”

I’m a seasoned genealogist, so I know unquestionably that my 23andMe Timeline is not only wrong, it’s entirely hogwash in 4 of 6 categories. A 5th category is close, and the 6th is reasonable but not verifiable.

The disparity of the British/Irish dates between 1759 when Henry Bolton was born in London and 1900-1930 is evident without discussion.  I do have a lot of British Isles ancestry, but it’s a result of many ancestors, not one and no one born there even remotely recently, let alone within the past generation. For me, someone born between 1900-1930 would be a parent.

Looking back at the Calculating Ethnicity Percentages article, you’ll note that I don’t have any Scandinavian ancestors in any known generation.  The 8% that 23andMe estimates, if accurate, equates to between a great-grandparent at 12.5% and a great-great-grandparent at 6.25%.  If the Scandinavian was one person, they would have been born in that timeframe (1750-1840) – but there was no one person.  The Scandinavian has to be very ancestral, meaning ancient Vikings or Normans or found in the Dutch population which is often found to be “Scandinavian.”  Regardless, there are no Scandinavian ancestors in my pedigree which reaches back well before 1750-1840.  Neither are there any Eastern European or Italian ancestors. None. Nada. Zip.

Perplexingly, it’s that unverifiable category, Native American, that so many people are desperately researching and scavenge for any possible clue.  There is no way to determine whether that category is right or wrong, so they will assume that it is accurate.  However, judging from the track record of the other categories – it’s more likely to be incorrect than correct.  Resorting to history alone, we know that the first European settlers arrived in North America in the early 1600s and my Native heritage is small, based on both my genealogy and my DNA, so a range of 1690-1790 would be a “good guess” with no genetic information at all.  My proven Native ancestors were born in the early/mid 1600s, but I have not successfully identified all of my Native ancestors, in particular the one(s) from my father’s side and when they were fully Native.

For a beginner or someone with unknown parentage, this timeline is horribly, horribly midleading and will cause novices to make massively incorrect assumptions. A British or Irish ancestor born between 1900-1930? Seriously?  This timeline combined with the 39.8% British/Irish suggests a parent.  Think about what an adoptee would take away from this timeline – and how their research could be derailed as a result.  Without parents available to DNA test, this erroneous information could make someone question their parentage.

Here’s an example of just how misleading this information can be.

In my case, I know beyond a doubt that my mother was primarily descended from German and Dutch recent immigrants with some French and Native American (Acadian) thrown in for good measure.  So, based on this timeline stating that a British/Irish ancestor was born in the British Isles between 1900 and 1930, combined with my ethnicity results of 39.8% British and Irish, OH MY GOD, my father is not who I thought, but is some British/Irish man.  MOTHER………………

All I can say is thank goodness I’ve done the DNA testing that I have and I know beyond a shadow of a doubt that my father is my biological father and not some British man, despite what this timeline suggests.  If I had no other evidence – I certainly would believe that my father was a British man, and I’d be GRATEFUL for this (highly erroneous) information.

On the flip side, many people will utilize this tool to “confirm” suspicions about genealogy.  I’ve already seen this happening on various lists.  With 4 of 6 categories being entirely, provably, incorrect, not to mention that the first category reflecting my largest percentage of ethnicity is so dramatically wrong, one can have absolutely no confidence in any of the other categories. I can’t and neither can anyone else.

I’m not alone either.  This, from another long-time genealogist: “I am dumbstruck.  It couldn’t be further from the truth for me.  I am very colonial on both sides.  Most recent immigrant ancestor was 1797.”  And from another: “No.  Just no.  Not accurate.”

So let me say this again.

You. Can. Have. No. Confidence.

If you already know your genealogy, then you don’t need this tool.  If you don’t know your genealogy, then you’re going to be misled by this tool.

It’s very clear that anyone with many ancestors that came from a particular population, but that haven’t been born in that location in many generations will have an incorrect timeline.  This would include just about everyone with colonial American roots.  The amount of a particular ethnicity does NOT equate to aggregating that ethnicity into a single ancestor and equating the amount of ethnicity to a recent birth in that location.  This logic is predicated on a whole lot of assumptions stacked on top of each other, like a house of cards. And we all know about assume.

23andMe, you should be ashamed of yourself for perpetrating genetic hogwash on your unsuspecting, believing and often vulnerable customers.  Climb down out of your ivory tower, buy a vowel and get a clue.  Statistics in an academic environment and reality sometimes just don’t mesh – and you, 23andMe, have the wherewithal and the customer base to discern the difference. You are supposed to be a science company.  You have no excuse.

I understand the desire to provide new tools to customers, but inaccurate simplicity is never a priority over realism.

I hope 23andMe will have the decency to remove this new deceptive and misleading “feature” that should never have made it past “proof of concept” in the first place.

thumbs-down

Concepts – Calculating Ethnicity Percentages

There has been a lot of discussion about ethnicity percentages within the genetic genealogy community recently, probably because of the number of people who have recently purchased DNA tests to discover “who they are.”

Testers want to know specifically if ethnicity percentages are right or wrong, and what those percentages should be. The next question, of course, is which vendor is the most accurate.

Up front, let me say that “your mileage may vary.” The vendor that is the most accurate for my German ancestry may not be the same vendor that is the most accurate for the British Isles or Native American. The vendor that is the most accurate overall for me may not be the most accurate for you. And the vendor that is the most accurate for me today, may no longer be the most accurate when another vendor upgrades their software tomorrow. There is no universal “most accurate.”

But then again, how does one judge “most accurate?” Is it just a feeling, or based on your preconceived idea of your ethnicity? Is it based on the results of one particular ethnicity, or something else?

As a genealogist, you have a very powerful tool to use to figure out the percentages that your ethnicity SHOULD BE. You don’t have to rely totally on any vendor. What is that tool? Your genealogy research!

I’d like to walk you through the process of determining what your own ethnicity percentages should be, or at least should be close to, barring any surprises.

By surprises, in this case, we’re assuming that all 64 of your GGGG-grandparents really ARE your GGGG-grandparents, or at least haven’t been proven otherwise. Even if one or two aren’t, that really only affects your results by 1.56% each. In the greater scheme of things, that’s trivial unless it’s that minority ancestor you’re desperately seeking.

A Little Math

First, let’s do a little very basic math. I promise, just a little. And it really is easy. In fact, I’ll just do it for you!

You have 64 great-great-great-great-grandparents.

Generation # You Have Who Approximate Percentage of Their DNA That You Have Today
1 You 100%
1 2 Parents 50%
2 4 Grandparents 25%
3 8 Great-grandparents 12.5%
4 16 Great-great-grandparents 6.25%
5 32 Great-great-great-grandparents 3.12%
6 64 Great-great-great-great-grandparents 1.56%

Each of those GGGG-grandparents contributed 1.56% of your DNA, roughly.

Why 1.56%?

Because 100% of your DNA divided by 64 GGGG-grandparents equals 1.56% of each of those GGGG-grandparents. That means you have roughly 1.56% of each of those GGGG-grandparents running in your veins.

OK, but why “roughly?”

We all know that we inherit 50% of each of our parents’ DNA.

So that means we receive half of the DNA of each ancestor that each parent received, right?

Well, um…no, not exactly.

Ancestral DNA isn’t divided exactly in half, by the “one for you and one for me” methodology. In fact, DNA is inherited in chunks, and often you receive all of a chunk of DNA from that parent, or none of it. Seldom do you receive exactly half of a chunk, or ancestral segment – but half is the AVERAGE.

Because we can’t tell exactly how much of any ancestor’s DNA we actually do receive, we have to use the average number, knowing full well we could have more than our 1.56% allocation of that particular ancestor’s DNA, or none that is discernable at current testing thresholds.

Furthermore, if that 1.56% is our elusive Native ancestor, but current technology can’t identify that ancestor’s DNA as Native, then our Native heritage melds into another category. That ancestor is still there, but we just can’t “see” them today.

So, the best we can do is to use the 1.56% number and know that it’s close. In other words, you’re not going to find that you carry 25% of a particular ancestor’s DNA that you’re supposed to carry 1.56% for. But you might have 3%, half of a percent, or none.

Your Pedigree Chart

To calculate your expected ethnicity percentages, you’ll want to work with a pedigree chart showing your 64 GGGG-grandparents. If you haven’t identified all 64 of your GGGG-grandparents – that’s alright – we can accommodate that. Work with what you do have – but accuracy about the ancestors you have identified is important.

I use RootsMagic, and in the RootsMagic software, I can display all 64 GGGG-grandparents by selecting all 4 of my grandparents one at a time.

In the first screen, below, my paternal grandfather is blue and my 16 GGGG-grandparents that are his ancestors are showing to the far right.  Please note that you can click on any of the images to enlarge.

ethnicity-pedigree

Next, my paternal grandmother

ethnicity-pedigree-1

Next, my maternal grandmother.

ethnicity-pedigree-2

And finally, my maternal grandfather.

ethnicity-pedigre-3

These displays are what you will work from to create your ethnicity table or chart.

Your Ethnicity Table

I simply displayed each of these 16 GGGG-grandparents and completed the following grid. I used a spreadsheet, but you can use a table or simply do this on a tablet of paper. Technology not required.

You’ll want 5 columns, as shown below.

  • Number 1-64, to make sure you don’t omit anyone
  • Name
  • Birth Location
  • 1.56% Source – meaning where in the world did the 1.56% of the DNA you received from them come from? This may not be the same as their birth location. For example an Irish man born in Virginia counts as an Irish man.
  • Ancestry – meaning if you don’t know positively where that ancestor is from, what do you know about them? For example, you might know that their father was German, but uncertain about the mother’s nationality.

My ethnicity table is shown below.

ethnicity-table

In some cases, I had to make decisions.

For example, I know that Daniel Miller’s father was a German immigrant, documented and proven. The family did not speak English. They were Brethren, a German religious sect that intermarried with other Brethren.  Marriage outside the church meant dismissal – so your children would not have been Brethren. Therefore, it would be extremely unlikely, based on both the language barrier and the Brethren religious customs for Daniel’s mother, Magdalena, to be anything other than German – plus, their children were Brethren..

We know that most people married people within their own group – partly because that is who they were exposed to, but also based on cultural norms and pressures. When it comes to immigrants and language, you married someone you could communicate with.

Filling in blanks another way, a local German man was likely the father of Eva Barbara Haering’s illegitmate child, born to Eva Barbara in her home village in Germany.

Obviously, there were exceptions, but they were just that, the exception. You’ll have to evaluate each of your 64 GGGG-grandparents individually.

Calculating Percentages

Next, we’re going to group locations together.

For example, I had a total of one plus that was British Isles. Three and a half, plus, that were Scottish. Nine and a half that were Dutch.

ethnicity-summary

You can’t do anything with the “plus” designation, but you can multiply by everything else.

So, for Scottish, 3 and a half (3.5) times 1.56% equals 5.46% total Scottish DNA. Follow this same procedure for every category you’re showing.

Do the same for “uncertain.”

Incorporating History

In my case, because all of my uncertain lines are on my father’s colonial side, and I do know locations and something about their spouses and/or the population found in the areas where each ancestor is located, I am making an “educated speculation” that these individuals are from the British Isles. These families didn’t speak German, or French, or have French or German, Dutch or Scandinavian surnames. People married others like themselves, in their communities and churches.

I want to be very clear about this. It’s not a SWAG (serious wild-a** guess), it’s educated speculation based on the history I do know.

I would suggest that there is a difference between “uncertain” and “unknown origin.” Unknown origin connotates that there is some evidence that the individual is NOT from the same background as their spouse, or they are from a highly mixed region, but we don’t know.

In my case, this leaves a total of 2 and a half that are of unknown origin, based on the other “half” that isn’t known of some lineages. For example, I know there are other Native lines and at least one African line, but I don’t know what percentage of which ancestor how far back. I can’t pinpoint the exact generation in which that lineage was “full” and not admixed.

I have multiple Native lines in my mother’s side in the Acadian population, but they are further back than 6 generations and the population is endogamous – so those ancestors sometimes appear more than once and in multiple Acadian lines – meaning I probably carry more of their DNA than I otherwise would. These situations are difficult to calculate mathematically, so just keep them in mind.

Given the circumstances based on what I do know, the 3.9% unknown origin is probably about right, and in this case, the unknown origin is likely at least part Native and/or African and probably some of each.

ethnicity-summary-2

The Testing Companies

It’s very difficult to compare apples to apples between testing companies, because they display and calculate ethnicity categories differently.

For example, Family Tree DNA’s regions are fairly succinct, with some overlap between regions, shown below.

ethnicity-ftdna-map

Some of Ancestry’s regions overlap by almost 100%, meaning that any area in a region could actually be a part of another region.

ethnicity-ancestry-map-2

For example look at the United Kingdom and Ireland. The United Kingdom region overlaps significantly into Europe.

ethnicity-ancestry-map

Here’s the Great Britain region close up, below, which is shown differently from the map above. The Great Britain region actually overlaps almost the entire western half of Europe.

ethnicity-ancestry-great-britain

That’s called hedging your bets, or maybe it’s simply the nature of ethnicity. Granted, the overlaps are a methodology for the vendor not to be “wrong,” but people and populations did and do migrate, and the British Isles was somewhat of a destination location.

This Germanic Tribes map, also from Ancestry’s Great Britain section, illustrates why ethnicity calculations are so difficult, especially in Europe and the British Isles.

ethnicity-invaders

Invaders and migrating groups brought their DNA.  Even if the invaders eventually left, their DNA often became resident in the host population.

The 23andMe map, below, is less detailed in terms of viewing how regions overlap.

ethnicity-23andme-map

The Genographic project breaks ethnicity down into 9 world regions which they indicate reflect both recent influences and ancient genetics dating from 500 to 10,000 years ago. I fall into 3 regions, shown by the shadowy Circles on the map, below.

ethnicity-geno-map-2

The following explanation is provided by the Genographic Project for how they calculate and explain the various regions, based on early European history.

ethnicity-geno-regions

Let’s look at how the vendors divide ethnicity and see what kind of comparisons we can make utilizing the ethnicity table we created that represents our known genealogy.

Family Tree DNA

MyOrigins results at Family Tree DNA show my ethnicity as:

ethnicity-ftdna-percents

I’ve reworked my ethnicity totals format to accommodate the vendor regions, creating the Ethnicity Totals Table, below. The “Genealogy %” column is the expected percentage based on my genealogy calculations. I have kept the “British Isles Inferred” percentage separate since it is the most speculative.

ethnicity-ftdna-table

I grouped the regions so that we can obtain a somewhat apples-to-apples comparison between vendor results, although that is clearly challenging based on the different vendor interpretations of the various regions.

Note the Scandinavian, which could potentially be a Viking remnant, but there would have had to be a whole boatload of Vikings, pardon the pun, or Viking is deeply inbedded in several population groups.

Ancestry

Ancestry reports my ethnicity as:

ethnicity-ancestry-amounts

Ancestry introduces Italy and Greece, which is news to me. However, if you remember, Ancestry’s Great Britain ethnicity circle reaches all the way down to include the top of Italy.

ethnicity-ancestry-table

Of all my expected genealogy regions, the most definitive are my Dutch, French and German. Many are recent immigrants from my mother’s side, removing any ambiguity about where they came from. There is very little speculation in this group, with the exception of one illegitimate German birth and two inferred German mothers.

23andMe

23andMe allows customers to change their ethnicity view along a range from speculative to conservative.

ethnicity-23andme-levels

Generally, genealogists utilize the speculative view, which provides the greatest regional variety and breakdown. The conservative view, in general, simply rolls the detail into larger regions and assigns a higher percentage to unknown.

I am showing the speculative view, below.

ethnicity-23andme-amounts

Adding the 23andMe column to my Ethnicity Totals Table, we show the following.

ethnicity-23andme-table-2

Genographic Project 2.0

I also tested through the Genographic project. Their results are much more general in nature.

ethnicity-geno-amounts

The Genographic Project results do not fit well with the others in terms of categorization. In order to include the Genographic ethnicity numbers, I’ve had to add the totals for several of the other groups together, in the gray bands below.

ethnicity-geno-table-2

Genographic Project results are the least like the others, and the most difficult to quantify relative to expected amounts of genealogy. Genealogically, they are certainly the least useful, although genealogy is not and never has been the Genographic focus.

I initially omitted this test from this article, but decided to include it for general interest. These four tests clearly illustrate the wide spectrum of results that a consumer can expect to receive relative to ethnicity.

What’s the Point?

Are you looking at the range of my expected ethnicity versus my ethnicity estimates from the these four entities and asking yourself, “what’s the point?”

That IS the point. These are all proprietary estimates for the same person – and look at the differences – especially compared to what we do know about my genealogy.

This exercise demonstrates how widely estimates can vary when compared against a relatively solid genealogy, especially on my mother’s side – and against other vendors. Not everyone has the benefit of having worked on their genealogy as long as I have. And no, in case you’re wondering, the genealogy is not wrong. Where there is doubt, I have reflected that in my expected ethnicity.

Here are the points I’d like to make about ethnicity estimates.

  • Ethnicity estimates are interesting and alluring.
  • Ethnicity estimates are highly entertaining.
  • Don’t marry them. They’re not dependable.
  • Create and utilize your ethnicity chart based on your known, proven genealogy which will provide a compass for unknown genealogy. For example, my German and Dutch lines are proven unquestionably, which means those percentages are firm and should match up relatively well to vendor ethnicity estimates for those regions.
  • Take all ethnicity estimates with a grain of salt.
  • Sometimes the shaker of salt.
  • Sometimes the entire lick of salt.
  • Ethnicity estimates make great cocktail party conversation.
  • If the results don’t make sense based on your known genealogical percentages, especially if your genealogy is well-researched and documented, understand the possibilities of why and when a healthy dose of skepticism is prudent. For example, if your DNA from a particular region exceeds the total of both of your parents for that region, something is amiss someplace – which is NOT to suggest that you are not your parents’ child.  If you’re not the child of one or both parents, assuming they have DNA tested, you won’t need ethnicity results to prove or even suggest that.
  • Ethnicity estimates are not facts beyond very high percentages, 25% and above. At that level, the ethnicity does exist, but the percentage may be in error.
  • Ethnicity estimates are generally accurate to the continent level, although not always at low levels. Note weasel word, “generally.”
  • We should all enjoy the results and utilize these estimates for their hints and clues.  For example, if you are an adoptee and you are 25% African, it’s likely that one of your grandparents was Africa, or two of your grandparents were roughly half African, or all four of your grandparents were one-fourth African.  Hints and clues, not gospel and not cast in concrete. Maybe cast in warm Jello.
  • Ethnicity estimates showing larger percentages probably hold a pearl of truth, but how big the pearl and the quality of the pearl is open for debate. The size and value of the pearl is directly related to the size of the percentage and the reference populations.
  • Unexpected results are perplexing. In the case of my unknown 8% to 12% Scandinavian – the Vikings may be to blame, or the reference populations, which are current populations, not historical populations – or some of each. My Scandinavian amounts translate into between 5 and 8 of my GGGG-grandparents being fully Scandinavian – and that’s extremely unlikely in the middle of Virginia in the 1700s.
  • There can be fairly large slices of completely unexplained ethnicity. For example, Scandinavia at 8-12% and even more perplexing, Italy and Greece. All I can say is that there must have been an awful lot of Vikings buried in the DNA of those other populations. But enough to aggregate, cumulatively, to between a great-grandparent at 12.5% and a great-great-grandparent at 6.25%? I’m not convinced. However, all three vendors found some Scandinavian – so something is afoot. Did they all use the same reference population data for Scandinavian? For the time being, the Scandinavian results remain a mystery.
  • There is no way to tell what is real and what is not. Meaning, do I really have some ancient Italian/Greek and more recent Scandinavian, or is this deep ancestry or a reference population issue? And can the lack of my proven Native and African ancestry be attributed to the same?
  • Proven ancestors beyond 6 generations, meaning Native lineages, disappear while undocumentable and tenuous ancestors beyond 6 generations appear – apparently, en masse. In my case, kind of like a naughty Scandinavian ancestral flash mob, taunting and tormenting me. Who are those people??? Are they real?
  • If the known/proven ethnicity percentages from Germany, Netherlands and France can be highly erroneous, what does that imply about the rest of the results? Especially within Europe? The accuracy issue is especially pronounced looking at the wide ranges of British Isles between vendors, versus my expected percentage, which is even higher, although the inferred British Isles could be partly erroneous – but not on this magnitude. Apparently part of by British Isles ancestry is being categorized as either or both Scandinavian or European.
  • Conversely, these estimates can and do miss positively genealogically proven minority ethnicity. By minority, I mean minority to the tester. In my case, African and Native that is proven in multiple lines – and not just by paper genealogy, but by Y and mtDNA haplogroups as well.
  • Vendors’ products and their estimates will change with time as this field matures and reference populations improve.
  • Some results may reflect the ancient history of the entire population, as indicated by the Genographic Project. In other words, if the entire German population is 30% Mediterranean, then your ancestors who descend from that population can be expected to be 30% Mediterranean too. Except I don’t show enough Mediterranean ancestry to be 30% of my German DNA, which would be about 8% – at least not as reported by any vendor other than the Genographic Project.
  • Not all vendors display below 1% where traces of minority admixture are sometimes found. If it’s hard to tell if 8-12% Scandinavian is real, it’s almost impossible to tell whether less than 1% of anything is real.  Having said that, I’d still like to see my trace amounts, especially at a continental level which tends to be more reliable, given that is where both my Native and African are found.
  • If the reason my Native and African ancestors aren’t showing is because their DNA was not passed on in subsequent generations, causing their DNA to effectively “wash out,” why didn’t that happen to Scandinavian?
  • Ethnicity estimates can never disprove that an ancestor a few generations back was or was not any particular ethnicity. (However, Y and mitochondrial DNA testing can.)
  • Absence of evidence is not evidence of absence, except in very recent generations – like 2 (grandparents at 25%), maybe 3 generations (great-grandparents at 12.5%).
  • Continental level estimates above 10-12 percent can probably be relied upon to suggest that the particular continental level ethnicity is present, but the percentage may not be accurate. Note the weasel wording here – “probably” – it’s here on purpose. Refer to Scandinavia, above – although that’s regional, not continental, but it’s a great example. My proven Native/African is nearly elusive and my mystery Scandinavian/Greek/Italian is present in far greater percentages than it should be, based upon proven genealogy.
  • Vendors, all vendors, struggle to separate ethnicity regions within continents, in particular, within Europe.
  • Don’t take your ethnicity results too seriously and don’t be trading in your lederhosen for kilts, or vice versa – especially not based on intra-continental results.
  • Don’t change your perception of who you are based on current ethnicity tests. Otherwise you’re going to feel like a chameleon if you test at multiple vendors.
  • Ethnicity estimates are not a short cut to or a replacement for discovering who you are based on sound genealogical research.
  • No vendor, NOT ANY VENDOR, can identify your Native American tribe. If they say or imply they can, RUN, with your money. Native DNA is more alike than different. Just because a vendor compares you to an individual from a particular tribe, and part of your DNA matches, does NOT mean your ancestors were members of or affiliated with that tribe. These three major vendors plus the Genographic Project don’t try to pull any of those shenanigans, but others do.
  • Genetic genealogy and specifically, ethnicity, is still a new field, a frontier.
  • Ethnicity estimates are not yet a mature technology as is aptly illustrated by the differences between vendors.
  • Ethnicity estimates are that. ESTIMATES.

If you like to learn more about ethnicity estimates and how they are calculated, you might want to read this article, Ethnicity Testing, A Conundrum.

Summary

This information is NOT a criticism of the vendors. Instead, this is a cautionary tale about correctly setting expectations for consumers who want to understand and interpret their results – and about how to use your own genealogy research to do so.

Not a day passes that I don’t receive very specific questions about the interpretation of ethnicity estimates. People want to know why their results are not what they expected, or why they have more of a particular geographic region listed than their two parents combined. Great questions!

This phenomenon is only going to increase with the popularity of DNA testing and the number of people who test to discover their identity as a result of highly visible ad campaigns.

So let me be very clear. No one can provide a specific interpretation. All we can do is explain how ethnicity estimates work – and that these results are estimates created utilizing different reference populations and proprietary software by each vendor.

Whether the results match each other or customer expectations, or not, these vendors are legitimate, as are the GedMatch ethnicity tools. Other vendors may be less so, and some are outright unethical, looking to exploit the unwary consumer, especially those looking for Native American heritage. If you’re interested in how to tell the difference between legitimate genetic information and a company utilizing pseudo-genetics to part you from your money, click here for a lecture by Dr. Jennifer Raff, especially about minutes 48-50.

Buyer beware, both in terms of purchasing DNA testing for ethnicity purposes to discover “who you are” and when internalizing and interpreting results.

The science just isn’t there yet for answers at the level most people seek.

My advice, in a nutshell: Stay with legitimate vendors. Enjoy your ethnicity results, but don’t take them too seriously without corroborating traditional genealogical evidence!

2016 Genetic Genealogy Retrospective

In past years, I’ve written a “best of” article about genetic genealogy happenings throughout the year. For several years, the genetic genealogy industry was relatively new, and there were lots of new tools being announced by the testing vendors and others as well.

This year is a bit different. I’ve noticed a leveling off – there have been very few announcements of new tools by vendors, with only a few exceptions.  I think genetic genealogy is maturing and has perhaps begun a new chapter.  Let’s take a look.

Vendors

Family Tree DNA

Family Tree DNA leads the pack this year with their new Phased Family Matches which utilizes close relatives, up to third cousins, to assign your matches to either maternal or paternal buckets, or both if the individual is related on both sides of your tree.

Both Buckets

They are the first and remain the only vendor to offer this kind of feature.

Phased FF2

Phased Family Matching is extremely useful in terms of identifying which side of your family tree your matches are from. This tool, in addition to Family Tree DNA’s nine other autosomal tools helps identify common ancestors by showing you who is related to whom.

Family Tree DNA has also added other features such as a revamped tree with the ability to connect DNA results to family members.  DNA results connected to the tree is the foundation for the new Phased Family Matching.

The new Ancient Origins feature, released in November, was developed collaboratively with Dr. Michael Hammer at the University of Arizona Hammer Lab.

Ancient European Origins is based on the full genome sequencing work now being performed in the academic realm on ancient remains. These European results fall into three primary groups of categories based on age and culture.  Customer’s DNA is compared to the ancient remains to determine how much of the customer’s European DNA came from which group.  This exciting new feature allows us to understand more about our ancestors, long before the advent of surnames and paper or parchment records. Ancient DNA is redefining what we know, or thought we knew, about population migration.

2016-ancient-origins

You can view Dr. Hammer’s presentation given at the Family Tree DNA Conference in conjunction with the announcement of the new Ancient Origins feature here.

Family Tree DNA maintains its leadership position among the three primary vendors relative to Y DNA testing, mtDNA testing and autosomal tools.

Ancestry

In May of 2016, Ancestry changed the chip utilized by their tests, removing about 300,000 of their previous 682,000 SNPs and replacing them with medically optimized SNPs. The rather immediate effect was that due to the chip incompatibility, Ancestry V2 test files created on the new chip cannot be uploaded to Family Tree DNA, but they can be uploaded to GedMatch.  Family Tree DNA is working on a resolution to this problem.

I tested on the new Ancestry V2 chip, and while there is a difference in how much matching DNA I share with my matches as compared to the V1 chip, it’s not as pronounced as I expected. There is no need for people who tested on the earlier chip to retest.

Unfortunately, Ancestry has remained steadfast in their refusal to implement a chromosome browser, instead focusing on sales by advertising the ethnicity “self-discovery” aspect of DNA testing.

Ancestry does have the largest autosomal data base but many people tested only for ethnicity, don’t have trees or have private trees.  In my case, about half of my matches fall into that category.

Ancestry maintains its leadership position relative to DNA tree matching, known as a Shared Ancestor Hint, identifying common ancestors in the trees of people whose DNA matches.

ancestry-common-ancestors

23andMe

23andMe struggled for most of the year to meet a November 2015 deadline, which is now more than a year past, to transition its customers to the 23andMe “New Experience” which includes a new customer interface. I was finally transitioned in September 2016, and the experience has been very frustrating and extremely disappointing, and that’s putting it mildly. Some customers, specifically international customers, are still not transitioned, nor is it clear if or when they will be.

I tested on the 23andMe older V3 chip as well as their newer V4 chip. After my transition to the New Experience, I compared the results of the two tests. The new security rules incorporated into the New Experience meant that I was only able to view about 25% of my matches (400 of 1651(V3) matches or 1700 (V4) matches). 23andMe has, in essence, relegated themselves into the non-player status for genetic genealogy, except perhaps for adoptees who need to swim in every pool – but only then as a last place candidate. And those adoptees had better pray that if they have a close match, that match falls into the 25% of their matches that are useful.

In December, 23andMe began providing segment information for ethnicity segments, except the parental phasing portion does not function accurately, calling into question the overall accuracy of the 23andme ethnicity information. Ironically, up until now, while 23andMe slipped in every other area, they had been viewed at the best, meaning most accurate, in terms of ethnicity estimates.

New Kids on the Block

MyHeritage

In May of 2016, MyHeritage began encouraging people who have tested at other vendors to upload their results. I was initially very hesitant, because aside from GedMatch that has a plethora of genetic genealogy tools, I have seen no benefit to the participant to upload their DNA anyplace, other than Family Tree DNA (available for V3 23andMe and V1 Ancestry only).

Any serious genealogist is going to test at least at Family Tree DNA and Ancestry, both, and upload to GedMatch. My Heritage was “just another upload site” with no tools, not even matching initially.

However, in September, MyHeritage implemented matching, although they have had a series of what I hope are “startup issues,” with numerous invalid matches, apparently resulting from their usage of imputation.

Imputation is when a vendor infers what they think your DNA will look like in regions where other vendors test, and your vendor doesn’t. The best example would be the 300,000 or so Ancestry locations that are unique to the Ancestry V2 chip. Imputation would result in a vendor “inferring” or imputing your results for these 300,000 locations based on…well, we don’t exactly know based on what. But we do know it cannot be accurate.  It’s not your DNA.

In the midst of this, in October, 23andMe announced on their forum that they had severed a previous business relationship with MyHeritage where 23andMe allowed customers to link to MyHeritage trees in lieu of having customer trees directly on the 23andMe site.  This approach had been problematic because customers are only allowed 250 individuals in their tree for free, and anything above that requires a MyHeritage subscription.  Currently 23andMe has no tree capability.

It appears that MyHeritage refined their DNA matching routines at least somewhat, because many of the bogus matches were gone in November when they announced that their beta was complete and that they were going to sell their own autosomal DNA tests. However, matching issues have not disappeared or been entirely resolved.

While Family Tree DNA’s lab will be processing the MyHeritage autosomal tests, the results will NOT be automatically placed in the Family Tree DNA data base.

MyHeritage will be doing their own matching within their own database. There are no comparison tools, tree matching or ethnicity estimates today, but My Heritage says they will develop a chromosome browser and ethnicity estimates. However, it is NOT clear whether these will be available for free to individuals who have transferred their results into MyHeritage or if they will only be available to people who tested through MyHeritage.

2016-myheritage-matches

For the record, I have 28 matches today at MyHeritage.

2016-myheritage-second-match

I found that my second closest match at MyHeritage is also at Ancestry.

2016-myheritage-at-ancestry

At MyHeritage, they report that I match this individual on a total of 64.1 cM, across 7 segments, with the largest segment being 14.9 cM.

Ancestry reports this same match at 8.3 cM total across 1 segment, which of course means that the longest segment is also 8.3 cM.

Ancestry estimates the relationship as 5th to 8th cousin, and MyHeritage estimates it as 2nd to 4th.

While I think Ancestry’s Timber strips out too much DNA, there is clearly a HUGE difference in the reported results and the majority of this issue likely lies with the MyHeritage DNA imputation and matching routines.

I uploaded my Family Tree DNA autosomal file to MyHeritage, so MyHeritage is imputing at least 300,000 SNPs for me – almost half of the SNPs needed to match to Ancestry files.  They are probably imputing that many for my match’s file too, so that we have an equal number of SNPs for comparison.  Combined, this would mean that my match and I are comparing 382,000 actual SNPs that we both tested, and roughly 600,000 SNPs that we did not test and were imputed.  No wonder the MyHeritage numbers are so “off.”

My Heritage has a long way to go before they are a real player in this arena. However, My Heritage has potential, as they have a large subscriber base in Europe, where we desperately need additional testers – so I’m hopeful that they can attract additional genealogists that are willing to test from areas that are under-represented to date.

My Heritage got off to a bit of a rocky start by requiring users to relinquish the rights to their DNA, but then changed their terms in May, according to Judy Russell’s blog.

All vendors can change their terms at any time, in a positive or negative direction, so I would strongly encourage all individuals considering utilizing any testing company or upload service to closely read all the legal language, including Terms and Conditions and any links found in the Terms and Conditions.

Please note that MyHeritage is a subscription genealogy site, similar to Ancestry.  MyHeritage also owns Geni.com.  One site, MyHeritage, allows individual trees and the other, Geni, embraces the “one world tree” model.  For a comparison of the two, check out Judy Russell’s articles, here and here.  Geni has also embraced DNA by allowing uploads from Family Tree DNA of Y, mitochondrial and autosomal, but the benefits and possible benefits are much less clear.

If the MyHeritage story sounds like a confusing soap opera, it is.  Let’s hope that 2017 brings both clarity and improvements.

Living DNA

Living DNA is a company out of the British Isles with a new test that purports to provide you with a breakdown of your ethnicity and the locations of your ancestral lines within 21 regions in the British Isles.  Truthfully, I’m very skeptical, but open minded.

They have had my kit for several weeks now, and testing has yet to begin.  I’ll write about the results when I receive them.  So far, I don’t know of anyone who has received results.

2016-living-dna

Genos

I debated whether or not I should include Genos, because they are not a test for genealogy and are medically focused. However, I am including them because they have launched a new model for genetic testing wherein your full exome is tested, you receive the results along with information on the SNPs where mutations are found. You can then choose to be involved with research programs in the future, if you wish, or not.

That’s a vastly different model that the current approach taken by 23andMe and Ancestry where you relinquish your rights to the sale of your DNA when you sign up to test.  I like this new approach with complete transparency, allowing the customer to decide the fate of their DNA. I wrote about the Genos test and the results, here.

Third Parties

Individuals sometimes create and introduce new tools to assist genealogists with genetic genealogy and analysis.

I have covered these extensively over the years.

GedMatch, WikiTree, DNAGedcom.com and Kitty Cooper’s tools remain my favorites.

I love Kitty’s Ancestor Chromosome Mapper which maps the segments identified with your ancestors on your chromosomes. I just love seeing which ancestors’ DNA I carry on which chromosomes.  Somehow, this makes me feel closer to them.  They’re not really gone, because they still exist in me and other descendants as well.

Roberta's ancestor map2

In order to use Kitty’s tool, you’ll have to have mapped at least some of your autosomal DNA to ancestors.

The Autosomal DNA Segment Analyzer written by Don Worth and available at DNAGedcom is still one of my favorite tools for quick, visual and easy to understand segment matching results.

ADSA Crumley cluster

GedMatch has offered a triangulation tool for some time now, but recently introduced a new Triangulation Groups tool.

2016-gedmatch-triangulation-groups

I have not utilized this tool extensively but it looks very interesting. Unfortunately, there is no explanation or help function available for what this tool is displaying or how to understand and interpret the results. Hopefully, that will be added soon, as I think it would be possible to misinterpret the output without educational material.

GedMatch also introduced their “Evil Twin” tool, which made me laugh when I saw the name.  Using parental phasing, you can phase your DNA to your parent or parents at GedMatch, creating kits that only have your mother’s half of your DNA, or your father’s half.  These phased kits allow you to see your matches that come from that parent, only.  However, the “Evil Twin” feature creates a kit made up of the DNA that you DIDN’T receive from that parent – so in essence it’s your other half, your evil twin – you know, that person who got blamed for everything you “didn’t do.”  In any case, this allows you to see the matches to the other half of your parent’s DNA that do not show up as your matches.

Truthfully, the Evil Twin tool is interesting, but since you have to have that parent’s DNA to phase against in the first place, it’s just as easy to look at your parent’s matches – at least for me.

Others offer unique tools that are a bit different.

DNAadoption.com offers tools, search and research techniques, especially for adoptees and those looking to identify a parent or grandparents, but perhaps even more important, they offer genetic genealogy classes including basic and introductory.

I send all adoptees in their direction, but I encourage everyone to utilize their classes.

WikiTree has continued to develop and enhance their DNA offerings.  While WikiTree is not a testing service nor do they offer autosomal data tools like Family Tree DNA and GedMatch, they do allow individuals to discover whether anyone in their ancestral line has tested their Y, mitochondrial or autosomal DNA.

Specifically, you can identify the haplogroup of any male or female ancestor if another individual from that direct lineage has tested and provided that information for that ancestor on WikiTree.  While I am generally not a fan of the “one world tree” types of implementations, I am a fan of WikiTree because of their far-sighted DNA comparisons, the fact that they actively engage their customers, they listen and they expend a significant amount of effort making sure they “get it right,” relative to DNA. Check out WikiTree’s article,  Putting DNA Results Into Action, for how to utilize their DNA Features.

2016-wikitree-peter-roberts

Thanks particularly to Chris Whitten at WikiTree and Peter Roberts for their tireless efforts.  WikiTree is the only vendor to offer the ability to discover the Y and mtDNA haplogroups of ancestors by searching trees.

All of the people creating the tools mentioned above, to the best of my knowledge, are primarily volunteers, although GedMatch does charge a small subscription service for their high end tools, including the triangulation and evil twin tools.  DNAGedcom does as well.  Wikitree generates some revenue for the site through ads on pages of non-members. DNAAdoption charges nominally for classes but they do have need-based scholarships. Kitty has a donation link on her website and all of these folks would gladly accept donations, I’m sure.  Websites and everything that goes along with them aren’t free.  Donations are a nice way to say thank you.

What Defined 2016

I have noticed two trends in the genetic genealogy industry in 2016, and they are intertwined – ethnicity and education.

First, there is an avalanche of new testers, many of whom are not genetic genealogists.

Why would one test if they weren’t a genetic genealogist?

The answer is simple…

Ethnicity.

Or more specifically, the targeted marketing of ethnicity.  Ethnicity testing looks like an easy, quick answer to a basic human question, and it sells kits.

Ethnicity

“Kim just wanted to know who she was.”

I have to tell you, these commercials absolutely make me CRINGE.

Yes, they do bring additional testers into the community, BUT carrying significantly misset expectations. If you’re wondering about WHY I would suggest that ethnicity results really cannot tell you “who you are,” check out this article about ethnicity estimates.

And yes, that’s what they are, estimates – very interesting estimates, but estimates just the same.  Estimates that provide important and valid hints and clues, but not definitive answers.

ESTIMATES.

Nothing more.

Estimates based on proprietary vendor algorithms that tend to be fairly accurate at the continental level, and not so much within continents – in particular, not terribly accurate within Europe. Not all of this can be laid a the vendor’s feet.  For example, DNA testing is illegal in France.  Not to mention, genetic genealogy and population genetics is still a new and emerging field.  We’re on the frontier, folks.

The ethnicity results one receives from the 3 major vendors (Ancestry, Family Tree DNA and 23andMe) and the various tools at GedMatch don’t and won’t agree – because they use different reference populations, different matching routines, etc.  Not to mention people and populations move around and have moved around.

The next thing that happens, after these people receive their results, is that we find them on the Facebook groups asking questions like, “Why doesn’t my full blooded Native American grandmother show up?” and “I just got my Ancestry results back. What do I do?”  They mean that question quite literally.

I’m not making fun of these people, or light of the situation. Their level of frustration and confusion is evident. I feel sorry for them…but the genetic genealogy community and the rest of us are left with applying ointment and Band-Aids.  Truthfully, we’re out-numbered.

Because of the expectations, people who test today don’t realize that genetic testing is a TOOL, it’s not an ANSWER. It’s only part of the story. Oh, and did I mention, ethnicity is only an ESTIMATE!!!

But an estimate isn’t what these folks are expecting. They are expecting “the answer,” their own personal answer, which is very, very unfortunate, because eventually they are either unhappy or blissfully unaware.

Many become unhappy because they perceive the results to be in error without understanding anything about the technology or what information can reasonably be delivered, or they swallow “the answer” lock stock and barrel, again, without understanding anything about the technology.

Ethnicity is fun, it isn’t “bad” but the results need to be evaluated in context with other information, such as Y and mitochondrial haplogroups, genealogical records and ethnicity results from the other major testing companies.

Fortunately, we can recruit some of the ethnicity testers to become genealogists, but that requires education and encouragement. Let’s hope that those DNA ethnicity results light the fires of curiosity and that we can fan those flames!

Education

The genetic genealogy community desperately needs educational resources, in part as a result of the avalanche of new testers – approximately 1 million a year, and that estimate may be low. Thankfully, we do have several education options – but we can always use more.  Unfortunately, the learning curve is rather steep.

My blog offers just shy of 800 articles, all key word searchable, but one has to first find the blog and want to search and learn, as opposed to being handed “the answer.”

Of course, the “Help” link is always a good place to start as are these articles, DNA Testing for Genealogy 101 and Autosomal DNA Testing 101.  These two articles should be “must reads” for everyone who has DNA tested, or wants to, for that matter.  Tips and Tricks for Contact Success is another article that is immensely helpful to people just beginning to reach out.

In order to address the need for basic understanding of autosomal DNA principles, tools and how to utilize them, I began the “Concepts” series in February 2016. To date I offer the following 15 articles about genetic genealogy concepts. To be clear, DNA testing is only the genetic part of genetic genealogy, the genealogical research part being the second half of the equation.

The Concepts Series

Concepts – How Your Autosomal DNA Identifies Your Ancestors

Concepts – Identical By Descent, State, Population and Chance

Concepts – CentiMorgans, SNPs and Pickin’ Crab

Concepts – Parental Phasing

Concepts – Y DNA Matching and Connecting With Your Paternal Ancestor

Concepts – Downloading Autosomal Data From Family Tree DNA

Concepts – Managing Autosomal DNA Matches – Step 1 – Assigning Parental Sides

Concepts – Genetic Distance

Concepts – Relationship Predictions

Concepts – Match Groups and Triangulation

Concepts – Sorting Spreadsheets for Autosomal DNA

Concepts – Managing Autosomal DNA Matches – Step 2 – Updating Matching Spreadsheets, Bucketed Family Finder Matches and Pileups

Concepts – Why DNA Testing the Oldest Family Members Is Critically Important

Concepts – Undocumented Adoptions Versus Untested Y Lines

My blog isn’t the only resource of course.

Kelly Wheaton provides 19 free lessons in her Beginners Guide to Genetic Genealogy.

Other blogs I highly recommend include:

Excellent books in print that should be in every genetic genealogist’s library:

And of course, the ISOGG Wiki.

Online Conference Resources

The good news and bad news is that I’m constantly seeing a genetic genealogy seminar, webinar or symposium hosted by a group someplace that is online, and often free. When I see names I recognize as being reputable, I am delighted that there is so much available to people who want to learn.

And for the record, I think that includes everyone. Even professional genetic genealogists watch these sessions, because you just never know what wonderful tidbit you’re going to pick up.  Learning, in this fast moving field, is an everyday event.

The bad news is that I can’t keep track of everything available, so I don’t mean to slight any resource.  Please feel free to post additional resources in the comments.

You would be hard pressed to find any genealogy conference, anyplace, today that didn’t include at least a few sessions about genetic genealogy. However, genetic genealogy has come of age and has its own dedicated conferences.

Dr. Maurice Gleeson, the gentleman who coordinates Genetic Genealogy Ireland films the sessions at the conference and then makes them available, for free, on YouTube. This link provides a list of the various sessions from 2016 and past years as well. Well worth your time!  A big thank you to Maurice!!!

The 19 video series from the I4GG Conference this fall is now available for $99. This series is an excellent opportunity for genetic genealogy education.

As always, I encourage project administrators to attend the Family Tree DNA International Conference on Genetic Genealogy. The sessions are not filmed, but the slides are made available after the conference, courtesy of the presenters and Family Tree DNA. You can view the presentations from 2015 and 2016 at this link.

Jennifer Zinck attended the conference and published her excellent notes here and here, if you want to read what she had to say about the sessions she attended. Thankfully, she can type much faster and more accurately than I can! Thank you so much Jennifer.

If you’d like to read about the unique lifetime achievement awards presented at the conference this year to Bennett Greenspan and Max Blankfeld, the founders of Family Tree DNA, click here. They were quite surprised!  This article also documents the history of genetic genealogy from the beginning – a walk down memory lane.

The 13th annual Family Tree DNA conference which will be held November 10-12, 2017 at the Hyatt Regency North Houston. Registration is always limited due to facility size, so mark your calendars now, watch for the announcement and be sure to register in time.

Summary

2016 has been an extremely busy year. I think my blog has had more views, more comments and by far, more questions, than ever before.

I’ve noticed that the membership in the ISOGG Facebook group, dedicated to genetic genealogy, has increased by about 50% in the past year, from roughly 8,000 members to just under 12,000. Other social media groups have been formed as well, some focused on specific aspects of genetic genealogy, such as specific surnames, adoption search, Native American or African American heritage and research.

The genetic aspect of genealogy has become “normal” today, with most genealogists not only accepting DNA testing, but embracing the various tools and what they can do for us in terms of understanding our ancestors, tracking them, and verifying that they are indeed who we think they are.

I may have to explain the three basic kinds of DNA testing and how they are used today, but no longer do I have to explain THAT DNA testing for genealogy exists and that it’s legitimate.

I hope that each of us can become an ambassador for genetic genealogy, encouraging others to test, with appropriate expectations, and helping to educate, enlighten and encourage. After all, the more people who test and are excited about the results, the better for everyone else.

Genetic genealogy is and can only be a collaborative team sport.

Here’s wishing you many new cousins and discoveries in 2017.

Happy New Year!!!

23andMe’s New Ancestry Composition (Ethnicity) Chromosome Segments

I was excited to see 23andMe’s latest feature that provides customers with Ancestry Composition (ethnicity) chromosome segment information by location.  This means I can compare my triangulation groups to these segments and potentially identify which ancestor’s DNA that I inherited carry which ethnicity – right?? Another potential way to help discern whether I should ask Santa for lederhosen or a kilt?

Not so fast…

Theoretically yes, but as it turns out, after working with the results, this tool doesn’t fulfill it’s potential and has some very significant issues, or maybe this new tool just unveiled underlying issues.

Rats, I guess Santa is off the hook.

Let’s take a look and step through the process.

Ancestry Composition Chromosome Painting

To see your Ancestry Composition ethnicity chromosome painting, sign into 23andMe, then go to the Reports tab at the top of your page and click on Ancestry. Please note that you can click on any of the graphics in this article to enlarge.

23andme-eth-seg-1

Then click on Ancestry Composition, which shows you the following:

23andme-eth-seg-2

Scrolling downs shows you your chromosomes, painted with your ethnicity. This isn’t new and it’s a great visual.

You may note that 23andMe paints both “sides” of each chromosome separately, the side you received from your mother and the side you received from your father. However, there is no way to determine which is which, and they are not necessarily the same side on each chromosome.

If one or both of your parents tested at 23andMe, you can connect your parents to your results and you can then see which ethnicity you received from which parent.

Let’s work through an example.

23andme-eth-seg-3

This person, we’ll call her Jasmine, received two segments of Native ancestry, one on chromsome 1 and one on chromosome 2, both on the first (top) strands or copies. She also received one segment of African on DNA strand (copy) 1 of chromsome 7.

Caveat

Words of warning.

JUST BECAUSE THESE ETNICITIES APPEAR ON THE SAME STRANDS OF DIFFERENT CHROMOSOMES, STRAND ONE IN THIS CASE, DOES NOT MEAN THEY ARE INHERITED FROM THE SAME PARENT.  

Each chromosome recombines separately and without a parent to compare to, there is no way to know which strand is mother’s or father’s on any chromsome. And figuring out which strand is which for one chromsome does NOT mean it’s the same for other chromsomes.

In fact, Jasmine’s mother has tested, and she has NO African on chromosome 7. However, Jasmine and her mother both have Native American on chromosomes 1 and 2 in the same location, so we know absolutely that Jasmine’s strand 1 on chromosome 7 is not from the same parent as strand 1 on chromosome 1 and 2, because Jasmine’s mother doesn’t have any African DNA in that location.

If you’re a seasoned 23andMe user, and you’re saying to yourself, “That’s not right, the chromosome sides should be aligned if a parent tests.”  You’re right, at least that’s what we’ve all thought.  Keep reading.

Let’s dig a bit further.

Connecting Up

23and Me encourages everyone to connect their parents, if your parents have tested.

Jasmine’s mother has tested and is connected to Jasmine at 23andMe.

23andme-eth-seg-4

Even though the button says “Connect Mother,” which makes it appear that Jasmine’s mother isn’t connected, she is. Clicking on Jasmine’s “Connect Mother” button shows the following:

23andme-eth-seg-5

Furthermore, if the parent isn’t connected, you don’t see any parental side ethnicity breakdown – and we clearly see those results for Jasmine.  Below is an example of the same page of someone whose parents aren’t connected – and you can see the verbiage at the bottom saying that a parent must be connected to see how much ancestry composition was inherited from each parent.

23andme-eth-seg-not-connect

If a child is connected to at least one parent, 23andMe, based on that parent’s test, tells the child which sides they inherited which pieces of their ethnicity from, shown for Jasmine, below.

23andme-eth-seg-6

In this case, the mother is connected to Jasmine and the father’s ethnicity results are imputed by subtracting the results where Jasmine matches her mother. The balance of Jasmine’s DNA ethnicity results that don’t match her mother in that location are clearly from her father.

23andMe may sort the results into the correct buckets, but they do not correctly rearrange the chromosome “copies” or “sides” on the chromosome browser display based on the parents’ DNA, as seen from the African example on chromosome 7. Either that, or the ethnicity phasing is inaccurate, or both.

You can see that 23andMe tells Jasmine that all of her Native is from her mother’s side, which is correct.

23andMe tells Jasmine that part of her North African and Sub-Saharan African are from her mother, but some North African is also from her father. You can see Jasmine’s African on her chromosome 7, below.

23andme-eth-seg-7

There is no African on Jasmine’s mother’s chromosome 7, below.

23andme-eth-seg-8

So if African exists on chromosome 7, it MUST come from Jasmine’s father’s side. Therefore, side one of chromosome 7 cannot be Jasmine’s mother’s side, because that’s where Jasmine’s African resides.

This indictes that either the results are incorrect, or the “sides” showing have not been corrected or realigned by 23andMe after parental ethnicity phasing, or both.

Here’s another example. Jasmine shows Middle East and North Africa on chromosomes 12 and 13 on sides one and two, respectively.

23andme-eth-seg-9

Jasmine’s mother shows Middle East and North Africa on chromosome 14, only, with none showing on chromosome 12 or 13.

23andme-eth-seg-10

Yet, 23andMe shows Jasmine receiving Middle East and North African DNA from her mother.

23andme-eth-seg-11

Jasmine is also shown as receiving Sub-Saharan African and West African from her mother, but Jasmine’s mother has no Sub-Saharan or West African, at all.

Interestingly, when you highlight both West African and Sub-Saharan African, shown below, it highlights the same segment of Jasmine’s DNA, so apparently these are not different categories, but subsets of each other, at least in this case, and reflect the same segment.

23andme-eth-seg-12

23andme-eth-seg-13

Jasmine’s mother shows this region of chromosome 7 to be “European” with no further breakdown.

Clearly Jasmine’s sides 1 and 2 have not been consistently assigned to her mother, because Jasmine’ African shows on both sides 1 and 2 of chromosomes 12 and 13 and Jasmine’s mother has no African on either on those chromosomes – so those segments should be assigned consistently to Jasmine’s father’s side, which, based on Jasmine’s match to her mother on chromosome 1, side 1 – Jasmine’s father’s “copy” should be Jasmine’s side 2.  This tool is not functioning correctly.

Jasmine’s father is deceased, so there is no way to test him.

The information provided by 23and Me contradicts itself.

Either the ethnicity assignment itself or the parental ethnicity phasing is inaccurate, or both. Additionally, we now know that the chromosome “sides,” meaning “copies” are inaccurately displayed, even when one parent’s DNA is available and connected, and the sides could and should be portrayed accurately.

This discrepancy has to be evident to 23andMe, if they are checking for consistency in assigning child to parent segments.  You can’t assign a child’s segment to a parent who doesn’t carry any of that ethnicity in a common location.  That situation should result in a big red neon sign flashing “STOP” in quality assurance.  Inaccurate results should never be delivered to testers, especially when there are easy ways to determine that something isn’t right.

The New Feature – Ethnicity Segments

Like I said, I was initially quite excited about this new feature, at least until I did the analysis. Now, I’m not excited at all, because if the results are flawed, so is the underlying segment data.

My original intention was to download the ethnicity segment information into my master spreadsheet so that I could potentially match the ethnicity segments against ancestors when I’ve identified an ancestral segment as belonging to a particular ancestral line.

This would have been an absolutely wonderful benefit.

Let’s walk though these steps so you can find your results and do your own analysis.

When you are on the Ancestry Composition page, you will be, by default, on the Summary page.

23andme-eth-seg-14

Click on the Scientific Details tab, at the top, and scroll down to the bottom of the page where you will see the following:

23andme-eth-seg-15

You will be able to select a confidence level, ranging from 50% to 90%, where 50% is speculative and 90% is the highest confidence. Hint – at the highest confidence level, many of the areas broken out in the speculative level are rolled up into general regions, like “European.”  Default is 50%.

23andme-eth-seg-16

Click on download raw data and you can then open or save a .csv file. I suggest then saving that file as an Excel file so you can do some comparisons without losing features like color.

In my case, I saved a 50% confidence file and a 90% confidence file to compare to each other.

I began my analysis with both strands of chromosome 1:

Strand 1 was easy.  (Click on graphic to enlarge.)

23andme-eth-seg-17

At the 50% confidence level, on the left, three segments are identified, but when you really look at the start and end positions, rows one and two overlap entirely. Looking back at the chromosome browser painting, this looks to be because that segment will show up in both of those categories, so this isn’t an either-or situation. Row 3 shows Scandinavian beginning at 79,380,466 and continuing through 230,560,900, which is a partial embedded segment of row 2.

At the 90% confidence level, on the right, above, this entire segment, meaning all of chromosome 1 on side 1, is simply called European.

You can see how this might get complex very quickly when trying to utilize this information in a Master DNA Spreadsheet with your matches, especially since individual segments can have 2 or 3 different labels.  However, I’d love to know where my mystery Scandinavian is coming from – assuming it’s real.

Now, let’s look at strand 2 of chromosome one. It’s a little more complex.

23andme-eth-seg-18

I’ve tried to color code identical, or partially-overlapping segments.

The red, green and apricot segments overlap or partially overlap at the 50% level, on the left, indicating that they show up in different categories.

The red segments are partially the same, with some overlapping, but are grouped differently within Europe.

The green Native/East Asian segments at the 90% level are interrupted by the blue unassigned segments in the middle of the green segments, while at the 50% confidence level, they remain contiguous.

All of the start and end segments change, even if the categories stay the same or generally the same. The grey example at the bottom is the easiest to see – the category changes to the more general “European” at the 90% level and the start segment is slightly different.

Jasmine and Her Mother

As one last example, let’s look at the segments at the 50% confidence level, which should be the least restrictive, that we were comparing when discussing Jasmine and her mother.

You can see, below, that Jasmine’s Native portion of chromosome 1 and 2 are either equal to or a subset of her mother’s Native portion, so these match accurately and are shown in green.

This tells us that Jasmine’s mother’s side of chromosomes 1 and 2 is Jasmine’s “copy 1” and given that we can identify Jasmine’s mother’s DNA, all of Jasmine’s “copy 1” should now be displayed as her mother’s DNA, but it isn’t.

23andme-eth-seg-19

On chromosomes 7 and 12, where Jasmine’s copy 1 shows African DNA, her mother has none. All African DNA segments are shown in red, above.

Furthermore, 23andMe attributes at least some portion of Jasmine’s African to Jasmine’s mother, but Jasmine’s mother’s only African DNA appears on chromosome 14, a location where Jasmine has none. There is no common African segment or segments between Jasmine and her mother, in spite of the fact that 23andMe indicates that Jasmine inherited part of her African DNA from her mother.  It’s true that Jasmine and her mother both carry African DNA, but not on any of the same segments, so Jasmine did not inherit her mother’s African DNA.  Jasmine’s African DNA had to have come from her father – and that’s evident if you compare Jasmine and her mother’s segment data.

Where Jasmine has African DNA segments, above, I’ve shown her mother’s corresponding DNA segments on both strands for comparison. I have not colored these segments. Conversely, where Jasmine’s mother has African, on chromosome 14, I have shown Jasmine’s corresponding DNA segments covering that segment.  There are no matches.

Clearly Jasmine did not inherit her African segments from her mother, or the segments have been incorrectly assigned as African or European, or multiple problems exist.

Summary

I initially thought the Ancestry Composition segments were a great addition to the genealogists toolset, but unfortunately, it has proven to be otherwise, highlighting deficiencies in more than one of the following area:

  • Potentially, the ancestry composition ethnicity breakdown itself.  Is the underlying ethnicity assignment incorrect?  In either case, that would not explain the balance of the issues we encountered.
  • The chromosome “sides” or “copy” shown after the parental phasing – in other words, the child’s chromosome copies can be assigned to a particular parent with either or both parents’ DNA. Therefore, after parental phasing, all of the same parent’s DNA should consistently be assigned to either copy 1 or copy 2 for the child on all of their chromosomes.  It isn’t.
  • The child’s ethnicity source (parent) assignment based on the parent’s or parents’ ethnicity assignment(s).  Hence, the African segment assignment issues above.
  • The ethnicity phasing itself.  The assigning of the source of Jasmine’s African DNA to her mother when they share no common African segments.  Clearly this is incorrect, calling into question the validity of the rest of the parental ethnicity phasing.

Unfortunately, we really don’t have adequate tools to determine exactly where the problem or problems lie, but problems clearly do exist. This is very disappointing.

As a result, I won’t be adding this information to my Master DNA spreadsheet, and I’m surely glad I took the time to do the analysis BEFORE I copied the segment data into my spreadsheet.  In my excitement, I almost skipped the analysis step, trusting that 23andMe had this right.

All ethnicity results need to be taken with a large grain of salt, especially at the intra-continent level, because the reference populations and technology just haven’t been perfected.  It’s very difficult to discern between countries and regions of Europe, for example.  I discussed this in the article, “Ethnicity Testing – A Conundrum.”

However, it appears that adding parental phasing on top means that instead of a grain of salt, we’re looking at the entire shaker, at least at 23andMe – even at the continent level – in this case, Africa, which should be easily discernable from European. Parental phasing by its very nature should be able to help refine our results, not make them less reliable.

Is this new segment information just showing us the problems with the original ethnicity information?  I hate to even think about this or ask these difficult questions, but we must, because testers often rely on minority (to them) ethnicity admixture information to help confirm the ethnicity of distant ancestors. Are the display tools or 23andMe’s programs not working correctly, or is there a deeper problem, or both?

I think I just received a big lump of coal, or maybe a chunk of salt, in my stocking for Christmas.

Bah, humbug.

23andMe V3 vs V4 Comparison

23andMe changed from their V3 to V4 kit back in November of 2013. At that time, they reduced the number of locations tested on their chip from over 900,000 by roughly one third to 577,382. This is not what would normally be considered an improvement. However, at that time, the FDA’s shutdown of their medical testing greatly overshadowed the new chip, so very little notice was taken in the community. Truthfully, many people didn’t expect them to survive, so the chip was a non-issue.

Frankly, I had hoped that with their medical testing revenue gone, that they would focus on what was left, genetic genealogy, but they didn’t. In fact, if anything, they turned their back on genetic genealogists with the new system redesign that was a result of the agreement with the FDA relative to what they can and cannot say and do, relative to reporting medical results.

Since it appears that 23andMe is going to survive, I wanted to see the difference between a 23andMe V3 kit and a V4 kit, both in terms of matching and ethnicity results. Plus, the new system experience is different than the old experience when I ordered more than 5 years ago, so I wanted to see what ordering today is like.

Ordering and Kit Registration

Ordering a second kit under the same account (mine) was just the beginning of the frustration. The system died when I was ordering, during the credit card portion, of course, and I had no idea if the order was placed or not. The order did go through, but my “order” stayed in the 23andMe shopping cart. However, my credit card was charged and the kit arrived.

Registering the kit was even more frustrating because I already had a user ID at 23andMe.

When you sign in to 23andMe, you must click Register Kit, then sign on, and not the reverse order, or your kit registration will be declined because you are already using your e-mail address at 23andMe. And no, there are no instructions for this…anyplace. And no, it’s not intuitive.

Terms of Service and Privacy Statements

I clicked through the Terms of Service and Privacy Statements so you can take a look, if you are so inclined. If you’re going to order, this is where the verbiage is contained that describes what 23andMe can do with your DNA in terms of sale to third parties and utilizing your DNA themselves to develop drugs and other products. Be sure you read each of these intertwined documents and understand what you are authorizing – even if you don’t “opt in” for research. There is a level of research that you cannot opt out of if you agree to the terms and conditions – and you can’t register your kit if you don’t agree to the terms and conditions.

v4-tos

Please note number 2 under “Privacy Highlights” above.

v4-tos2

Note number 4 under “Consent to the Use of Sensitive information” and the paragraph below number 4.

Here’s a link to the 23andMe Privacy Statement and the Terms of Service.

Next you see the Research Consent document.

v4-research-consent

Please note that every time you reply to one of their surveys, you are participating in research.  Please note that all of the people who have access to your information are not listed here, and you will have to refer to the Privacy Statement for complete information.

v4-research-consent-2

v4-research-consent-3

Please note that they can further analyze any saliva in a stored sample.

v4-research-consent-4

v4-research-consent-5

Please note that 23andMe reserves the right to develop products, and they have already filed for patents, the first of which was a for “designer baby” technology.  That was the day I withdrew my research consent.

v4-research-consent-6-2

Please note that even if you do not give consent, your information may still be used for other purposes, described in the Privacy Statement.  Furthermore, you cannot opt out of your aggregated (and anonymized) data being utilized.

v4-research-consent-7

What I didn’t see mentioned is that their processing is done by an outside company, LabCorp.

Lastly, the actual consent portion.

v4-research-consent-agree

Then you see this page, asking about sample storage:

v4-sample-storage

v4-submit

Last, you must click submit and you’re done with this part.

v4-complete

Well, you may think you’re done with registration, but 23andMe wastes no opportunity to attempt to collect information about you which makes your DNA information more valuable to the purchasers.

v4-survey

DNA Collection

Next, you have to spit in the vial, of course. A little vial doesn’t look like a lot of spit, until you’re trying to generate spit. Think of lemons. That helps.

I recently retested with both 23andMe and Ancestry, and while they both use a spit vial, the Ancestry vial is less messy because you don’t have to remove the top and screw on a lid.

The swab kits at Family Tree DNA are the best of all – no spitting needed.

Results

The results were available in about a month.

One of the reasons I order the V4 kit is because it seemed that almost everyone else had already transitioned to the New Experience, but I hadn’t. I wanted to see what the New Experience was all about. Wouldn’t you know it, not long after I ordered my V4 kit, my V3 kit finally transitioned to the “New Experience” and I discovered that the questions required for the New Experience transition are exactly the same as the V4 purchase. So in that sense, the experiences are now the same.

tne17

Unfortunately, you have re-authorize your participation in DNA Relatives when you transition to the “New Experience,” meaning that for the highly coveted comparison of DNA between your kit with multiple others, if your matches haven’t transitioned, or haven’t bothered to sign on and reauthorize, including open sharing – the previous sharing authorization isn’t sufficient. That’s sad, because a lot of people have gotten so disgusted with 23andMe that they don’t sign on anymore. Furthermore, if someone has for any other reason become disinterested or ill or died, their DNA is no longer working for anyone…except for 23andMe who continues to utilize their abandoned kit.

V3 Versus V4

I wanted to specifically compare two things between the V3 and the V4 kits – matches and ethnicity.

I wanted to know if the reduced number of SNPs on the V4 chip result in a smaller number of matches and I wanted to know if the ethnicity percentages changed between V3 and V4. I also wondered if there was any change between the old V3 ethnicity and the V3 New Experience ethnicity percentages.

Lastly, I wondered if there is any advantage for someone who has already taken the V3 test to order a V4 test, other than curiosity.

V3 “New Experience” Ethnicity

The ethnicity percentages remained exactly the same in the V3 old version and the V3 new experience version, which is exactly what I expected. The new display is shown below.

v3-ancestry

Where did the display on my chromosome go? Aha, it’s under Scientific Details – not what I would expect under that tab, but here it is.

v3-ancestry-chromosome

The bar above the browser shifts from Speculative to Conservative.

V4 Ethnicity

There is a very slight difference between the V3 and the V4 versions of ethnicity.

v4-ancestry

It looks like some of the unassigned results in the V3 version have become Native on chromosome 2 in the V4 version – by two tenths of one percent.

v4-ancestry-chromosome

Ethnicity Summary

The European remained the same, as did the Middle Eastern and North African, but the unassigned went from .2% to .1% and the difference is reflected in the East Asian and Native American that shifted from .3% to .5%.

In essence, there is very little difference.

V3 Versus V4 Matching

I set both accounts, meaning the V3 “new experience” and the V4 new account with exactly the same sharing options, meaning opting in to DNA Relatives and authorizing open sharing.

The V3 “new experience” which is the old test has a total of 1700 matches. The V4 kit on the newer chip with fewer SNPs has only 1651 matches, or 49 fewer, or about 3% of the matches. My assumption was that most of those matches were the same people on both chips.

However, when I compare my own result from the V3 and the V4 chip, 23andMe tells me that I have only 400 matches in common. Uh-oh. That’s vastly different than 1700 versus 1651.

v4-vs-v3-relatives

I scrolled down through this list of 400 people and I noticed that only the individuals who are coded as purple, meaning they are “open sharing” are listed. All of the people that gave me permission to share in V3 before are omitted. So all of that work, asking for permission back and forth for all those years, is apparently for naught unless they sign in and re-authorize DNA Relatives and authorize open sharing. What they previously authorized is irrelevant? Somehow, that just doesn’t seem right. Surely that can’t be true? Let’s see if we can figure this out.

I downloaded my aggregate match data for both the V3 “new experience” and V4 kit, color coded them, and combined them in one spreadsheet. Then I sorted by “Display Name.” Green rows = V3 file and red = V4 file.

v3v4-spreadsheet

I discovered some very interesting things.

  • The first thing I discovered is that 23andMe counts as a match anyone who matches you, whether or not you can see their chromosome data once their results are downloaded. See Jackson and Poole, above.
  • In the example above of my first 5 matches, you can see that for Jackson and Poole, their sharing choices for both V3 and V4 did not allow me to see their chromosome start or end locations, or number of SNPs. In other words, they were not sharing before and they aren’t sharing now. They are a grey (not sharing) or yellow (pending) dot on the V3 match list. On the V4 match list, anyone who has not authorized open sharing has a grey dot. For more discussion about the various dots and authorizations, see this article.
  • Halston is open sharing in both versions, indicated by a purple dot on the match list. I confirmed that he is open sharing in the new version. If you have reauthorized open sharing in either the New Experience, whether you were sharing or not before is irrelevant. Open sharing (or not) in the New Experience appears to override anything you did or didn’t authorize before.
  • Curtis is open sharing as well, but the end location and the number of SNPs is not the same between versions, although it is very close.
  • Davis is sharing in the old version (bluegreen dot on match list), but has not reauthorized in the V4 version, so I see that Davis does match me in both versions, above, but I can only see Davis’s chromosome information in V3 – the green rows, not in the V4 red row.  I can’t compare to other matches, called “Relatives in Common” in either version.  The information from the V3 test about “Relatives in Common” is shown below.

davis-sharing

This goes a long way in explaining why 23andMe says that I have 1700 (V3) and 1651 (V4) matches, but only 400 relatives in common. All of a sudden those 1700/1651 match numbers aren’t so terribly relevant or impressive anymore, because without the ability to see how I share, common surnames or a common ancestor in a tree, what is the point in even telling me that we match?

For anyone taking a V4 test, which is all that’s available now, those 400 people are the only ones I can see any chromosome information about at all.  So the other 1251 people, without any tree or chromosome information, are pretty much irrelevant.

There are three equally important parts of genetic genealogy:

  • That you match – at 23andMe, the list of DNA Relative matches (meaning the 1700/1651 matches)
  • How/where you match – at 23andMe, the chromosome information only available with open matching in V4 (meaning the 400 matches)
  • Genealogy information to go with the match – not available at 23andMe

In the V3 aggregated downloaded spreadsheet, there are a total of 1955 rows and in the V4 aggregated downloaded spreadsheet, there were a total of 1792 rows of matches. Many people match on more than one segment, which is why the discrepancy between the number of people who match and the rows of matches.

In the combined spreadsheet, there are a total of 170 rows that don’t appear in the other version. Of those 170, 153 are green, or are a V3 version match only and 17 are red, or the V4 version match only. I wouldn’t have expected any V4 only matches that are not also V3 matches. Unless 23andMe replaced some SNPs on the new chip, and didn’t just obsolete SNPs, I don’t know how one would explain these V4 only matches.

Of the 3747 total rows of matches, only 1421, or 42% actually have matching chromosome data listed. Of the 1421 total that have chromosome data, 1309 rows are listed in both the V3 and V4 versions.

1309 divided in two, for the green and red, would be about 654 in both red and green, which is getting close to the 400 common relatives that 23andMe reported. Several of those 654 rows do indeed have a second chromosome match row listed for the same person, so the 400 common relatives is certainly reasonable, working backwards, although it’s a LOT less than I expected from just looking at the total match numbers for V3 and V4 (1700/1651).

Last, it’s interesting to note for the close Denney relative shown below that not all of the start or end locations are the same, nor are the SNP counts. Furthermore, two locations aren’t listed at all in the other test. One segment on chromosome 9 of 8 cM from the red (V4) test is not listed in the V3 test results, and one segment on chromosome 13 of 16 cM in the green (V3) test is not listed in the V4 test results.

v3v4denne

While initially, the match number of either 1700 or 1651 seems enticing, things are not as they seemed on the surface.

  V3 Green V4 Red Other
Total Matches 1700 1651 49 total or 3% difference
Common Relatives 400 (23.5%) 400 (24%)
Rows in match Spreadsheet 1955 1792 163 difference
Color on Spreadsheet Green Red
Non-matching People to green or red 153 don’t match V4 17 don’t match V3 170 total
Anonymous 1166 total, or 31%
Rows with Chromosome data 654 match to red 654 match to green 1420 rows total, 112 that do not have a match between V3 and V4 spreadsheets

Why 1156 rows of matches, 31% to the total, found it necessary to list their name as “Anonymous” is beyond me.  The anonymous list runs on and on.

v3v4anonymous

Wouldn’t it just be easier to opt out of DNA Relatives? Perhaps in lieu of understanding that this is an option, they simply opted to list their name as anonymous. Unfortunately, they do still show as matches, inflating that match number but doing absolutely nothing else except adding an unnecessary level of frustration. All of the anonymous individuals are not sharing their chromosome data, which is to be expected, but also tells me that they figured out how not to share – so they could just as easily have opted out of DNA Relatives.

Why would 23andMe even show these people on my match list? Unfortunately, with the 2000 person cap, the anonymous people are all taking a match space, uselessly.  And 31% is not a trivial number of matches – in fact it’s a larger number, 512 of 1651 matches, than the number of matches, 400 of 1651, who have authorized open sharing.  That’s depressing.

Matching Summary

Before I had V4 match data to compare to V3 new experience data and realized that I only had 400 matches in common, meaning those who are have opted into DNA Relatives and are open sharing, despite having 1651 and 1700 “matches,” respectively, I really didn’t realize how abysmal the useful matching ratio at 23andMe really is.

Until I delved into what the various options really do and don’t mean, I didn’t realize that only 400 of my matches in the V4 version were sharing their DNA chromosome results with me and had authorized DNA Relatives.  That’s a very small percentage of the total.

As I did expect, for those matches found in one test and not the other, the V3 test is more productive than the V4 test, although the V4 test did have 17 matches that the V3 test did not. These 17 V4 only matches are hard to explain, because to the best of my knowledge, 23andMe did not replace any SNPs with other SNPs, although this is only logical answer that makes any sense for the V4 test matches that don’t have an equivalent V3 match – given that the V4 test is supposed to be a subset of V3. I suppose comparing the actual raw data file locations would answer that question, but I’m not motivated enough to do that. If someone else is, please add a comment to this article.

If my experience is representative of others, it would appear that people testing on the V3 chip will have more matches than the same person on the V4 chip, which makes sense because the number of testing locations is reduced on the V4 chip from about 950,000 on the V3 chip to about 577,000 on V4.

Three percent, or 49 total matches difference isn’t enough for me to worry about – especially not in light of the additional information that showed that my total useful matches weren’t 1700 or 1651, but in reality, 400 individuals in total, or about 25% of the total “matches.” If that 25% number holds on the 49 matches difference, that means that only about 12 of them would be even potentially useful.

The net-net of all of this is that there is absolutely no benefit to testing on the V4 chip if you have already tested on the V3 platform. There is no reason to order the V4 test.

The V3 test has more matches and included a much more robust health offering, so if you tested under V3, be glad that you tested when you did.

The End

With this article, and due to the reduced functionality for genealogists that I described in the “New 23andMe Experience – In a Word, Disappointing” article, I am officially finished dealing with 23andMe.  Only 400 matches who are fully sharing in the new version, with no genealogy information, just aren’t worth the associated headaches introduced by 23andMe – especially not when I have so many matches elsewhere who are interested in genealogy.

It didn’t have to be this way.  23andMe could have ported over the sharing authorizations for the V3 to the V3 New Experience and not required a new authorization set to participate in DNA Relatives with the people already previously authorized.  However, I suppose a new blanket authorization requirement for everyone covers the bases for 23andMe, but it leaves us genealogists out in the cold once again.  For me, that’s 325 people who were sharing in the old V3 version, but who have not authorized open sharing in the V3 New Experience, so I can’t see DNA Relatives in Common with those people now.  That’s almost as many people as who have authorized open sharing.

Both Family Tree DNA and Ancestry have significantly better products for the same or less money. Furthermore, 23andMe testers who are actually really interested in genealogy will have either uploaded their results to Family Tree DNA, retested at other vendors and/or uploaded their result to GedMatch. The rest of the people who tested at 23andMe aren’t interested in genealogy, think that their previous authorizations have them covered, or would probably be the 31% of my matches who list their name as anonymous.

If the 325 V3 matches who had previously authorized sharing were combined with the 400 open sharing matches, my total would be 725, which would be 44% of my matches.  Unfortunately, that’s not what 23andMe did.

Even though 23andMe clearly isn’t focused on genealogy, or genealogists, they apparently felt the financial heat of genealogists not placing orders after they raised their price to $199.

23andMe recently reverted to their $99 price for an “Ancestry Only” test, which omits the wellness and health factors. Ancestry.com’s DNA test is $99 and Family Tree DNA’s price for Family Finder is permanently reduced to $79 (or $39 to upload a 23andMe V3 file) and is paired with a lot fewer headaches, a much better user interface, trees and a focus on genealogy. I have more than enough matches at those two vendors, plus GedMatch, to keep me busy forever.

New 23andMe Experience – In a Word, Disappointing

Almost a year after the 23andMe “new experience” was promised “shortly” and then subsequently promised by 2015 year end, it’s finally here. In case you haven’t noticed, it’s September of 2016. I could have gestated a baby in less time. However, let’s take a look at the new experience process and features. I’m going to record each step in this new experience since I’ve finally transitioned.

Unfortunately, the new experience began with the 23andMe system either being very slow or not working at all, so I’ve pieced this together from several attempts over a couple of weeks. You’d think for as much as the new test costs, $199, twice that of their competitors and their own old test, they could at least have a reasonable system response time. If that happens as fast as the New Experience, it will be another year. I cannot even begin to tell you how many times I saw this screen.

tnebad

22andMe, you should be embarrassed. Really!

The “New 23andMe”

I thought the day would never arrive, but I did finally receive this e-mail:

tne1

Before we start reviewing the process and features, I want to mention that I did find an old to new feature converter, of sorts, provided by 23andMe. It’s not terribly useful, but it might be worth reviewing.

When you can get on and stay on the 23andMe system, you will see the following:

tne2

The system did not default to “me,” but to another one of the kits I manage. The next screen was to select primary profile.

My birth date was required. This is bothersome to me. It was never required before, and frankly, it’s none of their business. I answered it truthfully, only because I was afraid it would be part of a security question someplace down the line.

The next screen is shown below asking about your DNA Relatives Preferences. Apparently your old preferences don’t port to the new experience, at least not in total.

tne3

Here’s the infamous open sharing question that is supposed to replace all of the asking for permission to communicate and then asking for permission to share DNA segments. I say “supposed to,” because there is still a non-trivial amount of confusion surrounding options, as you’ll see shortly, but if you’re going to particulate in 23andMe for genealogy, do be sure to answer “yes.”

tne4

Here is what 23andMe has to say about the new open sharing option.

tne-sharing

Next, you can review your profile and verify, add to or change your information.

tne5

I personally think that displaying birth year is a potential security issue.

tne6

tne7

Next, 23andMe prompts you to compete a Health Profile.

The Health Profile started with a question marital status, which is again, none of their business. You can tell that their focus has really shifted to gathering information about you at every opportunity.

I’m not interested in providing them with any additional information they can then sell, so I’m not answering these questions.

tne8

You can opt instead to go to the home page, which is your new main account page, shown below.

tne9

You can see your account status information and the information available to you. All of the old functions have been redesigned, renamed or obsoleted. Figuring out which is which, and where, is like a scavenger hunt combined with a snipe hunt.

Ok, now you’re ready to begin looking around the new 23andMe site. I have a feeling that their earliest testers were some of the last to be converted, so if you’re already doing all of this, apologies. However, maybe you’ll learn something from my experiences or maybe you have something to add from your own!

Ancestry aka Ethnicity

Let’s start with Ancestry and the 3 reports 23andMe is showing. As a genealogist, I’m interested in the genealogy aspect of the 23andMe reports.

These are what we generally refer to as the ethnicity reports.

tne11

Let’s look first at Ancestry Composition

tne12

Where did the ethnicity display mapped onto my chromosome go? Aha, it’s under Scientific Details – not what I would expect under that tab, but here it is.

tne13

These colors are very difficult to distinguish from one another.

The bar above the browser shifts from Speculative to Conservative.

If you have a parent in the system, there used to be a “split view” where you could see your DNA “ancestry” as compared to that parent. That functionality is still there and is called “Inheritance View.”

I found the older “view” much easier to see and discern between the coloration. Here’s an example provided by 23andMe of the old versus the new.

tne-parent

Matches

Ok, let’s see if I can find my matches on this new system. Hmm, looking under tools, I see DNA Relatives, so I’ll click there. This used to be the Family Inheritance Advanced functionality.

I get to watch a tutorial first.

tne14

Looks like the new matching limit is 2000, a welcome increase. But why a match limit at all? Neither Family Tree DNA nor Ancestry have a match limit.

tne15

And of course the chromosome browser comparison. Interesting, they tell you THAT it’s available, but they don’t show you where to find this functionality. You’ll see that this becomes important later on.

tne16

Even though I’ve already opted into open sharing, I have to opt in again here and click on “View DNA Relatives.”

tne17

One thing that really bothers me is that after I clicked on “View DNA Relatives” as opposed to “I do not want to participate,” I could not go back or otherwise change that selection. I tried the settings option, by clicking on the profile name, and it appears that there is no option to rescind this permission.

DNA Relatives

Here is the list of my DNA Relatives. If you’re comparing this to a previous list, all of the information is missing on this page that was visible before, like haplogroups, genealogy surnames, etc., which made it easy to see at a glance.

There is however, a color coded sharing “dot” but with no legend, so I have NO IDEA who is sharing and who isn’t – or exactly what that means. Furthermore, I’m not colorblind, but the dot is so small (and I have 27 inch monitors) that I can’t tell if the dots are blue, green or some blue and some green – or maybe they are bluegreen.

After the fact, I stumbled on to the legend in the “sort by” box, but after reviewing the results, the legend makes no sense when seeing the sharing options and my cousins.

Let’s take a look.

tne18

So, the sharing legend is as follows:

  • Purple – open sharing
  • Blue/green – sharing
  • Yellow – Pending
  • Grey – not sharing

Let’s take a look at matches.

Blue Dot Match

According to the legend, a blue dot means sharing.

tne19

In order to see additional information, I click on my matches’ profile. Let’s start with my cousin Cheryl who has a blue dot.  I was sharing with Cheryl before the transition.

tne20

I can see my overlapping DNA with Cheryl, I can see her haplogroup and ethnicity, but at the bottom of the page, I cannot see any relatives in common because Cheryl has not participated in Open Sharing, according to the bottom of the screen shot below – although the blue/green dot indicates sharing, according to the legend. So does that mean we were sharing before (we were), but she has not clicked on open sharing since? And if so, what affect does that have? Which features and options are available under which kinds of old and new sharing combinations?  If Cheryl was sharing entirely with me before, which she was, why isn’t that sharing permission coming over into the new experience?  Why does she have to “reauthorize” sharing, if she has already given permission to share with me.  I’m confused, and let me say right here, that this question was never resolved.

tne21

On the right hand side of the page is a place to type a message and send to my match.

Fortunately, or unfortunately, whatever your perspective, my closest matches are people I know well and was sharing with before.  This does make it much easier to do comparisons between the old and new experiences.

Let’s check another blue dot cousin.

Blue Dot Match 2

The next cousin’s information that I checked invited me to take a look at his tree. Now, that’s interesting because I didn’t think that 23andMe had trees anymore, so I clicked on this link.

Aha, I can see his tree, but the message above the tree says this:

tne22

“As of May 1, 2015, the 23andMe Family Tree is view-only, and you are no longer able to edit or update your tree.  Your tree will remain available in this format in your account.  To edit or download your tree, import your tree data to MyHeritage.”

Of course, any tree with more than 250 people is not free at MyHeritage.

The match to this cousin says that he shows 103 surnames, but there is no matching surname feature to help me narrow down our matching surnames.

tne23

There should be no difference between this cousin’s sharing status and the first cousin, because they are both blue, and we were sharing before the transition, but I can’t see his “Relatives in Common” either.

So far, this is very discouraging, because I can’t do or see what I could before with the same people who have previously authorized sharing.  I know, in one case, that the person is no longer actively involved in genealogy and that means that I’ve lost functionality because they can’t or won’t “reauthorize” sharing.  Why should they need to?

Let’s move on.

Grey Dot Match

My third cousin has a grey dot and he is not participating in open sharing, so I can’t see his ancestry report, which I’m presuming here are my chromosome matches with him, or the Relatives in Common. Ironically, he had a profile message that says, “Just interested in learning more about my heritage and family history…”

Clearly he doesn’t understand the sharing options either.

Yellow Dot Match

Let’s try a cousin with a yellow sharing dot, which means pending, although I’m not sure exactly what is pending, where, and with whom.

tne24

Ok, this says that she is not sharing, in the top left corner, but that she has sent me a request to share Ancestry Reports. I’m open sharing, so why do I need to approve a request to share ancestry reports, and where do I do that?

23andMe does, however, show me our chromosome matches AND our relatives in common, even though we are supposedly “not sharing,” so I have no idea at all what else I would see if we were sharing.  In this case, what, exactly does “not sharing” mean and what else would I see by sharing?  Bizarre.

tne25

I notice that she has send me a message. Messages show in the right hand margin.  That’s a nice feature, but still not as nice as the ability to e-mail someone directly.

Purple Dot Match

Last, let’s try a cousin with the purple open sharing dot.

Well, this is really confusing, because it says that they are not sharing, but again, I can see our chromosome matches. That looks like sharing to me!  I clearly don’t understand what “not sharing” means.  It’s pretty much clear as mud.

tne26

I see a message at the bottom for me to request to share Ancestry Reports with her. However, I’m open sharing and since she has a purple dot, supposedly, so is she.

tne27

23andMe has really made a mess of “sharing,” both in terms of implementation, it appears, and assuredly in terms of explanation.  There is not one category of “sharing,” including when both people are open sharing in the new system, or when both people have previously authorized sharing in the old system, where I can see every category in the new system.

Chromosome Browser 5 Person Comparison

I spent a lot of time hunting for the ability to compare the 5 people in the chromosome browser, although minute by minute, I was quickly reaching the “I don’t care” point.

Under the DNA Relatives Tutorial, it clearly says you CAN compare up to 5 relatives, and this page says you can too, but where and how? 23andMe omitted a rather critical piece of information, it seems.

tne28

Please note that the above screen is displayed in Windows 10 using Internet Explorer, and if you scroll right, you can see more of the second column, but that’s all.

I finally found the Chromosome Browser that allows a comparison of up to 5 people, shown below. However, the function does not work correctly under Windows 10 with Internet Explorer. I switched to Edge and I could then see the compare option.  Believe it or not, it’s the same screen as above, but it doesn’t work correctly under Windows 10/Internet Explorer.

tne31

Half Versus Fully Identical Segments

Another feature that appears to have gone missing in the “New Experience” is the ability to see half versus fully identical segments.

Half siblings will have NO fully identical segments, because while they both inherited DNA from their common parent, the other parent was different, so no segments that they have should match at the same address on both chromosomes, meaning the chromosome they received from their mother and the chromosome they received from their father.

On the other hand, full siblings will have a non-trivial amount of fully identical segments, and this comparison was the easiest way to unquestionably tell a half from a full sibling. The previous version showed you segments that were half identical and fully identical, color coded.  The new version does not and only reports half identical segments.

When comparing my V3 test to my V4 test, 23andMe indicates that I am a “twin” to myself, so all of my segments should be fully identical when compared to myself, but looking at the comparison, only the half identical segments are reported now.

tne-half

Here’s an example (below) at GedMatch of the half versus full functionality.  The screen shot below shows my Ancestry V1 kit compared to my FTDNA kit.  You can see by the legend that the green bar indicates a full match and the yellow bar indicates a half match.  On chromosomes 1 and 2, which is all that I’ve shown, you can see the tiny sliver of yellow segments where one kit or the other doesn’t read the same address, so at that location, there is a mismatch of some sort.  At every “normal” location, I match myself fully because I’m my own “identical twin” as far as the system is concerned, and I share both parents DNA fully when compared to myself, so a “full match.”

tne-gedmatch

Furthermore, at 23andMe can you view the DNA comparison results in a table, but you can’t download them yet to a spreadsheet, although 23andMe indicates that this functionality is coming. However, it used to work.

Downloading Aggregate Data

At the bottom of the DNA Relatives page, I found the Download Aggregate Data button. The “Save As” did not work correctly under Windows 10/Internet Explorer, but I was able to open the file, then save it.

Share and Compare

I get to watch another tutorial. The Share and Compare function seems to be primarily for people who have immediate family who have tested, such as parents, grandparents or siblings.

tne29

The sharing and comparing all seems to be health except for Ancestry which is ethnicity. At the bottom, you can scroll through your matches and click on one to compare, and you’ll see much the same information as in the DNA Relatives section. If they are sharing health information, you’ll see more, such as traits.

Let’s see what else 23andMe has to offer.

Tools

On the Tools toolbar, I selected “All Tools.” We haven’t checked out “Family Tree” yet, so let’s do that. I didn’t think 23andMe had tree functionality anymore. Maybe this is a welcome surprise!

tne30

The Family Tree link takes you directly to MyHeritage. So no surprise, at least not a good one. Too bad.

Previous Health Reports

Because I tested prior to the 23andMe run-in with the FDA, my previous health reports are archived in the “Reports Archive.” I must say that the new traits are, for the most part, simply cocktail party conversation as compared to what we received before, and for half the price of current testing.

V3 testers do not receive the “Carrier Status” report, and this is the only test that is offered today that is actually medical in nature.

I would strongly suggest that anyone who actually wants health information test at either Ancestry.com for $99 or Family Tree DNA for $79 and then upload their results file to Promethease for $5. You’ll get a lot more than the very abbreviated 23andMe V4 information that costs $199.

Notice 23andMe doesn’t call the current product(s) health reports, but “wellness reports.” I think this is borderline deceptive except perhaps for Carrier Status.

tne32

Interestingly enough, both the Carrier Status and Traits reports under V4 require you to take an ethnicity survey before they show you your results, as does the Traits report under V3.

However, ethnicity is one of the things they are supposed to be telling you – in fact that’s one of the primary reasons people take these tests. So why do you have to tell them?

tne33

Download your Raw Data

Do download your raw data. You can upload it to GedMatch, to Promethease or depending on when you tested (after V2 and before V4, in November 2013) you can upload the file to Family Tree DNA for $39 in lieu of the $79 Family Finder test. The raw data download option is now under “Tools” on the toolbar.

tne35

You have to click on “I Understand” that you might discover sensitive health information about yourself or a family member.

tne36

On the first page below where you see the title “Your Raw Data,” click on the blue download button.

tne37

I encourage you to download your data while you are on the system, because it can be much, MUCH more difficult later, as I documented in this article.

Summary – Thumbs Down!!!!

As far as I’m concerned, there is nothing at 23andMe anymore for genealogists, especially when compared to the other testing companies, Family Tree DNA and Ancestry, who have both improved their offerings over the past several months.

23andMe provides fewer tools than they did previously to help genealogists identify their ancestors. As the other companies are making strides going forward, 23andMe is moving backwards.

23andMe doesn’t even provide anything as basic and simple as showing common surnames or a tree, both provided by Family Tree DNA and Ancestry. The new 23andMe interface is miserable and confusing, at best – for example – “sharing” which obviously doesn’t really mean sharing.  The new system is certainly not intuitive or written with a focus on genealogy, and their system times out horribly, outright fails and doesn’t work correctly with Internet Explorer on Windows 10. Many of the previous features used by genealogists have been obsoleted in this new version. Other than that, it’s wonderful (tongue firmly in cheek.)

As far as I’m concerned, genealogy testing at 23andMe is nothing more than a lure for 23andMe to obtain your DNA and answers to personal questions that are none of their business in order to utilize both for their own financial purposes.

Genealogists pulled 23andMe through the knothole by recommending them for testing when the FDA stopped 23andMe’s health testing. However, 23andMe, instead of enhancing their product for the genealogy market, has removed functionality, such as trees, Countries of Ancestry and full versus half identical segment identification – in essence stabbing genealogists in the back.

Both Family Tree DNA with their many tools and Ancestry, even without a chromosome browser, are both better choices. If it’s ethnicity testing you’re looking for, which is 23andMe’s strong point for genealogy, utilize either of the other vendors plus the many ethnicity (admixture) options at GedMatch.

The only person I would recommend 23andMe to now would be an adoptee looking for a very close match who did not find what they were looking for by testing with Family Tree DNA or Ancestry. In other words, I would only recommend 23andMe as a distant third and only in a pinch. For the normal genealogist, the other two vendors’ data bases and tools have become so large and robust that there just isn’t any reason to test at 23andMe.

I will continue to periodically check the 23andMe site, not for genealogy, but because I believe my father had additional children and I still have hopes of finding them or their children. I wish that 23andMe had implemented an option for notification of “immediate or close family” matches, but then again, they would have to be focused on genealogy in order to do that.

I have written one more article comparing the 23andMe V3 versus the V4 test matching and ethnicity, which holds some real surprises, but aside from publishing that article and an occasional check for my father’s possible offspring, I’m done with 23andMe, completely, entirely, finit, kaput, forever. I didn’t even bother to integrate my match file again in my DNA Master Spreadsheet. Downloading data with no corresponding ability to contact the tester (aside from the 23andMe message system on a website not functioning property), with an extremely low response rate, no trees and not even matching surnames isn’t fun, it’s simply frustrating.

23andMe is now far more work than pleasure and I’m simply done with them. As far as I’m concerned, they’ve had 3 years now to get their act together since the FDA issue…and they haven’t. The “new experience” has gotten worse, not better. The only positive aspect of the new experience is the new limit of 2000 matches, compared to no limit at the other vendors, open sharing, although there is still confusion surrounding that, and the fact that multiple profiles are now managed separately – thankfully. The other vendors have never been this unnecessarily complex relative to open sharing or multiple accounts, so they don’t have a corresponding mess to unravel.

There is a great irony here, because with 23andMe being the first vendor in the autosomal marketspace that was commercially viable could have owned the show, but they’ve blown it, over and over again. And they just blew it one last time.

I give the 23andMe “new experience” a big thumbs down.

How Much DNA Do We Share? It Depends

I was curious how testing the same two people at the 3 different vendors, then uploading the results from those different vendors to GedMatch and repeating the matching process there would affect the amount of DNA reported as matching.

I have a third cousin who has tested at all 3 labs independently, meaning they did not upload a file from either 23andMe or Ancestry to Family Tree DNA. Furthermore, they downloaded their 23andMe and Family Tree DNA files to GedMatch. They have not downloaded their Ancestry results to GedMatch, so I can’t do the Ancestry to Ancestry comparison, unfortunately.

So, we have one pair of third cousins, 3 individual vendor tests (each) and 8 independent answers to the question, “How much DNA do we share?”.

First, the theoretical expected average (as reported on the ISOGG wiki page) is 53 cM for third cousins. Blaine Bettinger’s actual findings through the shared cM project indicate an average of 79 cM for third cousins, and the actual range found is 0-198 cM, after removing outliers. This isn’t the first time in genetic genealogy that we’ve found that the theoretical or expected results aren’t what really happens as we learn more about how DNA actually works.

Let’s see how reality stacks up for our third cousin pair.

Vendor Threshold Total cM Total Segments Largest Segment Est Relationship
Theoretical 3C Average, Actual Average and Actual Range 53 ISOGG, 79 Actual, Range(0-198)
At Vendors
FTDNA 7cM/500 SNPs 149*** 22 33.52 2nd-3rd cousin
23andMe 7cM/700 SNPs 134 6 40.8 2nd-3rd cousin
Ancestry V1 5cM after Timber** 132 8 Not provided 3rd-4th cousin
At GedMatch
GedMatch 1* (23andMe V3 to 23andMe V3) 7cM/700 SNP 147 6 43.7 3.3 gen to MRCA****
GedMatch 2* (FTDNA to FTDNA) 7cM/700 SNP 136 6 43.7 3.4 gen to MRCA****
GedMatch 3* (23andMe V3 to FTDNA) 7cM/700 SNP 136 6 43.7 3.4 gen to MRCA****
GedMatch 4* (Ancestry V1 to 23andMe V3) 7cM/700 SNPs 147.5 6 43.7 3.3 gen to MRCA****
GedMatch 5* (Ancestry V1 to FTDNA) 7cM/700 SNPs 147.5 6 43.7 3.3 gen to MRCA****

Total cM is rounded except for 147.5, which doesn’t round in either direction.

*GedMatch at default setting which is currently 7cM and 700 SNPs.

**Unknown if SNPs are being utilized at Ancestry as a threshold parameter, and if so, the threshold is unknown.

***Total cM at Family Tree DNA includes small segments if you match. At 23andMe and GedMatch, total segments means only the total number of segments over the match threshold. The number at Family Tree DNA would be 112 cM if only counting segments greater than 5cM and 107 if only counting cM greater than 7. Of note, in my comparison, there no matching segments between 5.48 and 11.09, so this may be an unusual circumstance.

****The actual generations to a common recent ancestor (MRCA) is 4, counting our parents as generation 1.  It is unclear whether GedMatch counts you as generation 1 or your parents as generation 1.

Results like this are a perfect illustration of why relationship ranges based on DNA are ranges, not absolutes. I know, unquestionably that my cousin is my third cousin. However, were I to utilize ONLY the averages, I would be looking at either a 2nd cousin utilizing the theoretical numbers or a 2nd cousin once removed utilizing the real average, neither of which are accurate in this case.  Averages are made up of everyone in the range, smallest to largest – and in this case, the results fall into the larger than average category.

All of the Total cM numbers are two to three times the theoretical expected Total cM, but all of the Total cMs are still within the observed and reported range for third cousins.

For more on relationship ranges, theoretical expected versus actual and ranges as reported from crowd sourced information see here and here and here.

Blaine Bettinger provides a free download of his latest Shared cM Project results, which includes a great chart on the last page that provides a minimum, average and max cM shown for each relationship type. Thanks Blaine, for this very useful tool!