mtDNA Haplogroup K Project Cladogram
as of 14 June 2007
The cladogram published here is somewhat different from the previous phylogenetic diagrams I have published. It shows the various subclades of mtDNA haplogroup K found in the K Project in proportion to their numbers in the Project. The nodes have the subclade names, not individual membersí kit numbers. Only the defining mutations for the subclades were used to create the cladogram; the mutations are not shown. I will not explain here which mutations are required for all the subclades; I did that in my last MitoSearch survey. Only the 305 high-resolution (HVR1 + HVR2) FTDNA sequences in the Project were used. Subclade assignments were based on the 43 members with confirmed subclades from full-sequence test results plus my experience in predicting subclades from HVR results. The numbers for subclade membership range from one in several subclades to 71 in K1a+. Subclades with one member are usually from single full-sequence tests. For further information on the definition of the subclades Dr. Doron Beharís K tree (Figure 1) should be consulted. The cladogram was created using Tom Gladís mtDNAtool and Fluxus-Engineering Network software, for which links may be found in the survey mentioned above.
The beginning of the chart is at the top with the basic K. There are no members with just the basic mutations. The complete list of the ancestral K mutations is in MitoSearch at arfhh.
To the right of K the first node is
The first small node to the left of K is K1. There is only one example, based on a full-sequence test. This subclade has the defining coding-region mutations for K1, but not the required mutations for K1a, K1b or K1c. The 5000-year-old ÷tzi the Iceman was in this subclade.
K1a+ is the largest group for two reasons. One, K1a in general is over 60% of K. Two, many of the lower subclades of K1a are defined by coding-region mutations.
Working through K1aís lower subclades from left to right, K1a4a1 is the surprise produced by our full-sequence tests. We already have seven members in this subclade, which is defined by coding-region mutations, while many of its sister K1a subclades (K1a2, K1a5, K1a6, K1a7 and K1a8) have not shown up at all. Thatís probably due to the difference in the populations used by Behar for his tree and that of the FTDNA customers. A few exact HVR matches have been included for this cladogram, although one other exact match is not a K1a4a1. Note that there are no examples of the upstream K1a4 or K1a4a subclades.
K1a11 is my provisional designation for a cluster which has three confirmed members in the Project, with three more matches. It is easily identified by several HVR mutations, 16T being the key one due to its rarity. This subclade is probably the single unlabeled example on Beharís tree between K1a9 and K1a1, although that example has mutation 17 rather than 16. I suspect a typo somewhere.
The large group marked 195C+ is also labeled with just that mutation on Beharís tree. It includes the two clusters I have elsewhere called Pre-K1a10 and Pre-K1a9, based on whether the sequences have position 524 insertions or not. There is some guesswork involved in these assignments. One predicted example turned out to be a K1a4a1. Those with the additional 114T have been assigned instead to K1a1. K1a9 is the second largest Ashkenazi subclade. K1a10 is my provisional term for the very large group based on 16048A. The node here represents 24 individuals; Behar only found one in the population he used. K1a9 and K1a10 each have three confirmed examples. The ones Iíve called K1a10, as well as those in K1a11 above, are listed in the Project database as simply K1a* until a new ďofficialĒ tree is released.
K1a1 has one confirmed example to which I added three more with 114T. There are also single confirmed examples of K1a1a and K1a1b. K1a1b1 has two confirmed examples to which I have added a few more possible examples. K1a1b1a is the largest Ashkenazi subclade. The Project has six confirmed examples; I have added several more almost certainly in that subclade. The whole K1a1 group has many examples which might move up- or downstream if full-sequence tests were taken.
K1a3a has one confirmed example. I have not tried to predict others for this subclade, but there are potential members listed under K1a+.
K1b+ includes those probably in K1b, but not in the easily identified K1a1a or K1b2 subclades. Those two lower subclades have one and two confirmed examples, to which I have added others based on matching HVR results.
K1c+ is probably restricted to unconfirmed members of K1c1, which has only one confirmed member so far. K1c2 has three confirmed members and many others, since it is easily determined from HVR results.
Over the months and years the nodes with plus signs should decrease in size as more K Project members take full-sequence tests and may then be assigned to the more specific subclades. It should then become easier to predict subclades from exact or very close matches. Some of the new test results will fall into subclades without current representation. Also, new subclades will emerge as did K1a10 and K1a11.
Those who have been determined to be in mtDNA Haplogroup K may and should join our K Project by clicking on the blue Join button on their FTDNA personal pages. Further information may be found on our website.
William R. Hurst
Administrator, mtDNA Haplogroup K Project