42 Full-Sequence mtDNA Results and
Subclades
From Haplogroup K Project
May 24, 2007
This CHART,
as discussed below, represents the current 42 members, or 10%, of the mtDNA
Haplogroup K Project at FamilyTreeDNA who have received results from
full-sequence mtDNA tests. Two more tests are in progress. I published similar
charts when there were eight and 42 test results available. The chart was
produced using Fluxus-Engineering Network software, using both
the Reduced Median and Median Joining algorithms. The input file was created
using Tom Glad’s mtDNAtool.
On the
chart, KROOT represents the ancestral K haplotype, as shown in this MitoSearch entry;
so all mutations on the chart occurred since the founding of K. With two
exceptions, I have used all of the HVR mutations on each individual’s K Project
entry. Instead of using one, two or three pairs of the position 524 insertions
I have used “524i” to simplify the chart. Also, to save room I omitted the
relatively insignificant 309.1C and 309.2C insertions. However, the
coding-region (CR) mutations used were not
from the individuals’ full-sequence test results, since I don’t have access to
all of those and some may contain medically-relevant information. Instead, I
have used only the coding-region mutations which define the various assigned
subclades from Dr. Doron Behar’s 2006 paper which
contains the current “official” K tree. For this chart I follow the most recent
practice of FTDNA to not use
asterisks; so I marked K1a* and K2a* as simply K1a and K2a. This time I have
tried to mark all back mutations. However, I have not put a subclade label on
every node; if there is no label on a node, it will be found at the previous
branching point.
From KROOT,
the line to the right is one of the two major divisions of K,
Mutations
1189 and 10398 define the other major K division, K1. New to this chart is the
addition of sequence #46547 designated as just K1, since it does not have the
defining mutations for any of the three major branches of K1. Besides that one,
the three major subclades of K1 diverge from branching point mv8. The first
shown is K1c, defined by 498- and two other HVR mutations. There are no examples
of a plain K1c; if one existed it would be at mv9. K1c2 is defined by 16320T
and three CR mutations. Three examples are shown. K1c1 is defined only by CR
mutations; there is only one example so far.
K1b is
defined by the CR mutation 5913. My substitute term “524i” for the
length-heteroplasmic 524 insertions makes its first appearance here. Although
the three examples here have these insertions, in reality they only appear in
about half of K1b haplotypes. K1b2, defined by one HVR and two CR mutations,
has two different haplotypes here. K1b1 is defined by three CR mutations, but
here there is only one example of its lower subclade K1b1a. That’s defined by
three more CR mutations and two HVR ones, mainly 16319A. Most of these,
including the current example, also have 16463G. No examples of K1b1b or K1b1c
have been tested in the Project.
K1a,
defined solely by HVR2 mutation 497T, is about 60% of K. Here it is somewhat
overrepresented at 71%. Branching off to the right from mv4 is the largest
group K1a1, defined by 11914. There is one example of this subclade as well as
one K1a1a, defined by one more CR mutation. The larger branch is that of K1a1b,
defined by 15924. One example of the basic subclade has four additional HVR
mutations. K1a1b1 is defined by adding 11470. Two separate examples based on
Behar’s tree are shown, one of which has picked up the interesting mutation
114T. Beyond mv7 are six examples of the largest Ashkenazi subclade K1a1b1a,
which is defined by two CR mutations and 16234T. Behar shows 114T below 16234T;
it just happens that these five examples all have it. 16223T usually occurs in
a higher percentage of this subclade than the one-of-six represented here.
Plain K1a
is more or less a grab-bag of K1a haplotypes which don’t have the mutations to
be in lower subclades K1a1 through K1a9. There are eight of those here. Five of
them have 195C, a group, including K1a9, which includes about 18% of the Ks on
MitoSearch as opposed to the 8.5% on Behar’s tree. I think the difference is
the larger representation of British Isles and especially
Up from
mv4, defined by four CR mutations is the large subclade K1a4a1, represented by
seven different haplotypes, five with the 524 insertions, two without. This is
the largest subclade which can’t easily be predicted from HVR results, since it
has no defining HVR mutations. Two examples have 16261T; the jury is still out
on its significance. Many sequences in K1a with 524 insertions, but not 195C,
may be candidates for K1a4a1, as are those with 16245T.
A new
subclade has shown up recently, to the right and down from mv4. K1a3a is defined
by four CR mutations and represented by #78430. No examples have tested so far
from subclade K1a2 or from K1a5 to K1a8. All those are defined only by CR
mutations.
In general,
the 42 tests are a fair representation of K as shown on Behar’s tree. The
exceptions were discussed above. There are missing subclades, but consider that
Behar’s tree was based on 121 full-sequence samples.
Anyone who has
tested as a K is invited to join the K Project by clicking on the blue Join
button on your FTDNA personal page. Additional information is available at our website.
©2007
William R. Hurst
Administrator, mtDNA Haplogroup K Project
