Site icon DNAeXplained – Genetic Genealogy

Family Tree DNA’s Mitochondrial Haplotree

Roberta Estes

On September 27th, 2018 Family Tree DNA published the largest Y haplotree in the world, based on SNP tests taken by customers. Now, less than two weeks later, they’ve added an exhaustive mitochondrial DNA (mtDNA) public haplotree as well, making this information universally available to everyone.

Family Tree DNA’s mtDNA Haplotree is based on the latest version of the mtDNA Phylotree. The new Family Tree DNA tree includes 5,434 branches derived from more than 150,000 full sequence results from 180+ different countries of origin. Family Tree DNA‘s tree has SIX TIMES more samples than the Phylotree. Furthermore, Family Tree DNA only includes full sequence results, where Phylotree includes partial results.

This new tree is a goldmine! What does it provide that that’s unique? Locations – lots of locations!

The Official Phylotree

Unlike the Y DNA tree, which is literally defined and constructed by the genetic community, new mitochondrial DNA branches cannot be added to the official mitochondrial Phylotree by Family Tree DNA. Haplogroups, meaning new branches in the form of SNPs are added to the Y tree as new SNPs are discovered and inserted into the tree in their proper location. The mitochondrial DNA phylotree can’t be expanded by a vendor in that manner.

The official mitochondrial Phylotree is maintained at www.phylotree.org and is episodically updated. The most recent version was mtDNA tree build 17, published and updated in February 2016. You can view version history here.

Mitochondrial Phylogenic Tree Version 17

Version 17 of the official mitochondrial tree consists of approximately 5,400 nodes, or branches with a total of 24,275 samples uploaded by both private individuals and academic researchers which are then utilized to define haplogroup branches.

Individuals can upload their own full sequence results from Family Tree DNA, but they must be in a specific format. I keep meaning to write detailed instructions about how to submit your full sequence test results, but so far, that has repeatedly slipped off of the schedule. I’ll try to do this soon.

In a nutshell, download your FASTA file from Family Tree DNA and continue with the submission process here. The instructions are below the submission box, so scroll down.

In any case, the way that new branches are added to the phylotree is when enough new results with a specific mutation are submitted and evaluated, the tree will have a new branch added in the next version. That magic number of individuals with the same mutation was 3 in the past, but now that so many more people are testing, I’m not sure if that number holds, or if it should. Spontaneous mutations can and do happen at the same location. The Phylotree branches mean that the haplogroup defining mutations indicate a common ancestor, not de novo separate mutations. That’s why analysis has to be completed on each candidate branch.

How do Mitochondrial DNA Branches Work?

If you are a member of haplogroup J1c2f today, and a certain number of people in that haplogroup have another common mutation, that new mutation may be assigned the designation of 1, as in J1c2f1, where anyone in haplogroup J1c2f who has that mutation will be assigned to J1c2f1.

While the alternating letter/number format is very easy to follow, some problems and challenges do exist with the alternating letter/number haplogroup naming system.

The Name of the Game

The letter number system works fine if not many new branches are added, branches don’t shuffle and if the growth is slow. However, that’s not the case anymore.

If you recall, back in July of 2012, which is equivalent to the genetic dark ages (I know, right), the Y tree was also represented with the same type of letter number terminology used on the mitochondrial tree today.

For example, Y DNA haplogroup R-M269 was known as R1b1a2, and before that the same haplogroup was known as R1b1c. The changes occurred because so many new haplgroups were being discovered that a new sprout wasn’t added from time to time, but entire branches had to be sawed off and either discarded or grafted elsewhere. It became obvious that while the R1b1a2 version was nice, because it was visually obvious that R1b1a2a was just one step below R1b1a2, that long term, that format just wasn’t going to be able to work anymore. New branches weren’t just sprouting, wholesale shuffling was occurring. Believe it or not, we’re still on the frontier of genetic science.

In 2012, the change to the SNP based haplogroup designations was introduced by Family Tree DNA, and adopted within the community.

The ISOGG tree, the only tree that still includes the older letter/number system and creates extended letter number haplogroup names as new SNPs are added provides us with an example of how much the Y tree has grown.

You can see that the letter/number format haplogroups to the far right are 19 locations in length. The assigned SNP or SNPs associated with that haplogroup are shown as well. Those 19-digit haplogroup names are just too unwieldy, and new haplogroups are still being discovered daily.

It’s 2012 All Over Again

That’s where we are with mitochondrial DNA today, but unlike Y DNA naming, a vendor can’t just make that change to a terminal SNP based naming system because all vendors conform to the published Phylotree.

However, in this case, the vendor, Family Tree DNA has more than 6 times the number of full sequence mitochondrial results than the mitochondrial reference model Phylotree. If you look at the haplogroup projects at Family Tree DNA, you’ll notice that (some) administrators routinely group results by a specific mutation that is found within a named haplogroup, meaning that the people with the mutation form a subgroup that they believe is worthy of its own haplogroup subgroup name. The problem is that unless enough people upload their results to Phylotree, that subgroup will never be identified, so a new haplogroup won’t be added.

If the entire Family Tree DNA data base were to be uploaded to Phylotree, can you imagine how many new haplogroups would need to be formed? Of course, Family Tree DNA can’t do that, but individual testers can and should.

Challenges for Vendors

The challenge for vendors is that every time the phylotree tree is updated and a new version is produced, the vendors must “rerun” their existing tester samples against the new haplogroup defining mutations to update their testers’ haplogroup results.

In some cases, entire haplogroups are obsoleted and branches moved, so it’s not a simple matter of just adding a single letter or digit. Rearranging occurs, and will occur more and more, the more tests that are uploaded to Phylotree.

For example, in the Phylotree V17 update, haplogroup A4a1 became A1a. In other words, some haplogroups became entirely obsolete and were inserted onto other branches of the tree.

In the current version of the Phylotree, haplogroup A4 has been retired.

Keep in mind that all haplogroup assignments are the cumulative combination of all of the upstream direct haplogroups. That means that haplogroup A4a1, in the prior version, had all of the haplogroup defining mutations shown in bold in the chart below. In the V17 version, haplogroup A1a contains all of the mutations shown in bold red. You might notice that the haplogroup A4 defining mutation T16362C is no longer included, and haplogroup A4, plus all 9 downstream haplogroups which were previously dependent on T16362C have been retired. A4a1 is now A1a.

Taking a look at the mitochondrial tree in pedigree fashion, we can see haplogroup A4a1 in Build 15 from September 2012, below.

Followed by haplogroup A1a in the current Build 17.

Full Sequence Versus Chip Based Mitochondrial Testing

While Family Tree DNA tests the full sequence of their customers who purchase that level of testing, other vendors don’t, and these changes wreak havoc for those vendors, and for compatibility for customer attempting to compare between data bases and information from different vendors.

That means that without knowing which version of Phylotree a vendor currently uses, you may not be able to compare meaningfully with another user, depending on changes that occurred that haplogroup between versions. You also need to know which vendor each person utilized for testing and if that vendor’s mitochondrial results are generated from an autosomal style chip or are actually a full mitochondrial sequence test. Utilizing the ISOGG mtDNA testing comparison chart, here’s a cheat sheet.

Vendor No Mitochondrial Chip based haplogroup only mitochondrial Full Sequence mitochondrial
Family Tree DNA No Yes – V17
23andMe Yes – Build V7 No
Ancestry None
LivingDNA Yes – Build V17 No
MyHeritage None
Genographic V2 Yes – Build V16 No

Of the chip-based vendors, 23andMe is the most out of date, with V7 extending back to November of 2009. The Genographic Project has done the best job of updating from previous versions. LivingDNA entered the marketplace in 2016, utilizing V17 when they began.

Family Tree DNA’s mitochondrial test is not autosomal chip based, so they don’t encounter the problem of not having tested needed locations because they test all locations. They have upgraded their customers several times over the years, with the current version being V17.

Family Tree DNA’s mitochondrial DNA test is a separate test from their Family Finder autosomal test while the chip-based vendors provide a base-level haplogroup designation that is included in their autosomal product. However, for chip-based vendors, updating that information can be very challenging, especially when significant branch changes occur.

Let’s take a closer look.

Challenges for Autosomal Chip-Based Vendors Providing Mitochondrial Results

SNP based mitochondrial and Y DNA testing for basic haplogroups that some vendors include with autosomal DNA is a mixed blessing. The up side, you receive a basic haplogroup. The down aide, the vendor doesn’t test anyplace near all of the 16,569 mitochondrial DNA SNP locations.

I wrote in detail about how this works in the article, Haplogroup Comparisons Between Family Tree DNA and 23andMe. Since that time, LivingDNA has also added some level of haplogroup reporting through autosomal testing.

How does this work?

Let’s say that a vendor tests approximately 4000 mitochondrial DNA SNPs on the autosomal chip that you submit for autosomal DNA testing. First, that’s 4000 locations they can’t use for autosomal SNPs, because a DNA chip has a finite number of locations that can be utilized.

Secondly, and more importantly, it’s devilishly difficult to “predict” haplogroups at a detailed level correctly. Therefore, some customers receive a partial haplogroup, such as J1c, and some receive more detail.

It’s even more difficult, sometimes impossible, to update haplogroups when new Phylotree versions are released.

Why is Haplogroup Prediction and Updating so Difficult?

The full mitochondrial DNA sequence is 16,569 locations in length, plus or minus insertions and deletions. The full sequence test does exactly what that name implies, tests every single location.

Now, let’s say, by way of example, that location 10,000 isn’t used to determine any haplogroup today, so the chip-based vendors don’t test it. They only have room for 4000 of those locations on their chip, so they must use them wisely. They aren’t about to waste one of those 4000 spaces on a location that isn’t utilized in haplogroup determination.

Let’s say in the next release, V2, that location 10,000 is now used for just one haplogroup definition, but the haplogroup assignment still works without it. In other words, previously to define that haplogroup, location 9000 was used, and now a specific value at location 10,000 has been added. Assuming you have the correct value at 9,000, you’re still golden, even if the vendor doesn’t test location 10,000. No problem.

However, in V3, now there are new haplogroup subgroups in two different branches that use location 10,000 as a terminal SNP. A terminal SNP is the last SNP in line that define your results most granularly. In haplogroup J1c2f, the SNP(s) that define the f are my terminal SNPs. But if the vendor doesn’t test location 10,000, then the mutation there can’t be used to determine my terminal SNP, and my full haplogroup will be incomplete. What now?

If location 10,000 isn’t tested, the vendor can’t assign those new haplogroups, and if any other haplogroup branch is dependent on this SNP location, they can’t be assigned correctly either. Changes between releases are cumulative, so the more new releases, the further behind the haplogroup designations become.

Multiple problems exist:

The lack of complete testing also means that while Family Tree DNA can publish this type of tree and contribute to science, the other vendors can’t.

Let’s take a look at Family Tree DNA’s new tree.

Finding the Tree

To view the tree, click here, but do NOT sign in to your account. Simply scroll to the bottom of the page where you will see the options for both the Y DNA Haplotree and the mtDNA Haplotree under the Community heading.

Click on mtDNA Haplotree.

If you are a Family Tree DNA customer, you can view both the Y and mitochondrial trees from your personal page as well. You don’t have to have taken either the Y or mitochondrial DNA tests to view the trees.

Browsing the mtDNA Tree

Across the top, you’ll see the major haplogroups.

I’m using haplogroup M as an example, because it’s far up the tree and has lots of subgroups. Only full sequence results are shown on the tree.

The basic functionality of the new mitochondrial tree, meaning how it works, is the same as the Y tree, which I wrote about in the Family Tree DNA’s PUBLIC Y DNA Haplotree.

You can view the tree in two formats, countries or variants, in the upper left-hand corner. View is not the same thing as search.

When viewing the mitochondrial DNA phylotree by country, we see that haplogroup M has a total of 1339 entries, which means M and everything below M on the tree.

However, the flags showing in the M row are only for people whose full mitochondrial sequence puts them into M directly, with no subgroup.

As you can see, there are only 12: 6 people in Australia, and one in 5 other countries. These are the locations of the most distant known ancestor of those testers. If they have not completed the maternal Country of Origin on the Earliest Known Ancestor tab, nothing shows for the location.

Viewing the tree by variant shows the haplogroup defining mutations, but NOT any individual mutations beyond those that are haplogroup defining.

For each haplogroup, click on the three dots to the right to display the country report for that haplogroup.

The Country Report

The Country Report provides three columns.

The column titled Branch Participants M shows only the total of people in haplogroup M itself, with no upstream or downstream results, meaning excluding M1, M2, etc. Just the individuals in M itself. Be sure to note that there may be multiple pages to click through, at bottom right.

The second column, Downstream Participants – M and Downstream (Excluding other Letters) means the people in haplogroup M and M subclades. You may wonder why this column is included, but realize that branches of haplogroup M include haplogroups G, Q, C, Z, D and E. The middle column only includes M and subgroups that begin with M, without the others, meaning M, M10, M11 but not G, Q, etc.

Of course the final column, All Downstream Participants – M and Downstream (Including other Letters) shows all of the haplogroup M participants, meaning M and all subclades, including all other haplogroups beneath M, such as M10, G, Q, etc..

What Can I Do with This Information?

Unlike the companion Y tree DNA, since surnames change every generation for maternal lineages, there is no requirement to have multiple matching surnames on a branch to be displayed.

Therefore, every person who includes a location for a most distant known ancestor is included in the tree, but surnames are not.

I want to see, at a glance, where the other people in my haplogroup, and the haplogroups that are the “direct ancestral line” of mine are found today. Clusters may mean something genealogically or are at least historically important – and I’ll never be able to view that information any other way. In fact, before this tree was published, I wasn’t able to see this at all. Way to go Family Tree DNA!!

It’s very unlikely that I’ll match every person in my haplogroup – but the history of that haplogroup and all of the participants in that haplogroup are important to that historical lineage of my family. At one time, these people all shared one ancestor and determining when and where that person lived is relevant to my family story.

Searching for Your Haplogroup

I’m searching for haplogroup J1c2f by entering J1c2f in the “Go to Branch Name.”

There it is.

I can see that there are 17 people in Sweden, 13 in Norway, 5 in Germany, 3 in Russia, etc. What’s with the Scandinavian cluster? My most distant known ancestor was found in Germany. There’s something to be learned here that existing records can’t tell me!

The mother branch is J1c2 which shows the majority of individuals in Ireland followed by England. This probably suggests that while J1c2f may have been born in Scandinavia, J1c2 probably was not. According to the supplement to Dr. Doron Behar’s paper, A “Copernican” Reassessment of the Human Mitochondrial DNA tree from its Root, which provides ages for some mitochondrial DNA haplogroups:

Haplogroup How Old Standard Deviation Approximate Age Range in Years
J1c2 9762 2010 7,752 – 11,772
J1c2f 1926 3128 500 – 5,054

I happen to know from communicating with my matches that the haplogroup J1c2f was born more than 500 years ago because my Scandinavian mito-cousins know where their J1c2f cousin was then, and so do I. Mine was in Germany, so we know our common ancestor existed sometime before that 500 year window, and based on our mutations and the mutation tree we created, probably substantially before that 500 year threshold.

Given that J1c2, which doesn’t appear to have been born in Scandinavia is at least 7,700 years old, we can pretty safely conclude that my ancestor wasn’t in Scandinavia roughly 9,000 years ago, but was perhaps 2,000 years, ago when J1c2f was born. What types of population migration and movement happened between 2,000 and 9,000 years ago which would have potentially been responsible for the migration of a people from someplace in Europe into Scandinavia.

The first hint might be that in the Nordic Bronze Age, trade with European cultures became evident, which of course means that traders themselves were present. Scandinavian petroglyphs dating from that era depict ships and art works from as far away as Greece and Egypt have been found.

The climate in Scandinavia was warm during this period, but later deteriorated, pushing the Germanic tribes southward into continental Europe about 3000 years ago. Scandinavian influence was found in eastern Europe, and numerous Germanic tribes claimed Scandinavian origins 2000 years ago, including the Bergundians, Goths, Heruls and Lombards.

Hmmm, that might also explain how my mitochondrial DNA, in the form of my most distant known ancestor arrived in Germany, as well as the distribution into Poland.

Is this my family history? I don’t know for sure, but I do know that the clustering information on the new phylotree provides me with clustering data to direct my search for a historical connection.

What Can You Do?

The next few steps aren’t related to actually having your results displayed on the phylotree, but they are important to taking full advantage of the power of testing.

Of course, always click on save when you’re finished.

These steps assure that you’ve maximized the benefits of your mitochondrial results for your own research and to your matches as well. Collaborative effort in completing geographic and known ancestor information means that we can all make discoveries.

The article, Working with Mitochondrial DNA Results steps you through you all of the various tools provided to Family Tree DNA testers.

Now, go and see who you match, where your closest matches cluster, and on the new mtDNA Haplotree, what kind of historical ancestral history your locations may reveal. What’s waiting for you?

______________________________________________________________

Disclosure

I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

Genealogy Services

Genealogy Research

Exit mobile version