[Note: This forms part of a longer
conversation, I have extraced it because I feel it is an
important contribution. W.H.S.D.]
In his
recent email, Ed asked a lot of questions. First, if you don't
quite understand the difference between an STR count mutation
and a SNP mutation, you should read at least the first page of
the attached
tutorial which I wrote 6 years ago when BigY first appeared.
Some things have changed in the 6 years since, like the
reference genome for the Y-chromosome, which means the location
identifier for SNP mutations probably has changed, but it does
not change the basics which is all we need to answer the
questions. I am going to try to answer all of them by inserting
them here, in italics, followed by my answer and comments.
Bill Hough wrote: "Back to Douglass. We have one BigY result,
Duncan's, which with all the STR test matches, proves that
William was not among the Scottish ancestors of a lot of other
Douglas testers. Duncan shares this SNP with three other
subjects, none named Douglas. SNP results from another
descendant of William might produce SNPs that are common with
some of Duncan's currently unique SNPs, which when combined with
the paper trees might serve as proof of descent from our
William. As I said, you know the results when you see the
results."
Would you expand on this, Bill?
I screwed up the first sentence. I meant the descendants of
William 1610, and therefore William himself, are not among
descendants of the Scottish ancestors of many of the other
Douglas participants.
Look at the FTDNA Douglas project page to start. That is here:
https://www.familytreedna.com/public/Douglas/default.aspx?section=yresults
.
That page gives the results of STR testing of men who have the
Douglas surname and wondered how they are related to each other,
so they purchased a test from FTDNA or another lab. The numbers
in the column opposite a participant's kit number are the counts
of repeats in short sequences of repeats of the DNA chemical
bases at STR marker locations. The names of the markers are at
the top of each of the columns. To my knowledge, the word
"marker" is only used to refer to these locations where Short
Tandem Repeats occur. An STR mutation is a change in counts at
one of these locations (markers) from one generation to the
next. Such mutations are relatively frequent and can be in
either direction, an increase or a decrease in the count of
repeats.
A SNP is an actual change in a chemical base at a specific
position on the chromosome. The only information on SNPs that
appear on the page we are looking at is the column labeled
Haplogroup, and the description of each group of STR results. In
the case of the descendants of William, that is the sequence
(R-U106>L48>Z8>Z343) given it the header of this group.
R1b Group Type 1: (R-U106>L48>Z8>Z343). From an ancestor of
William Douglas, 1610, New London
This is a sequence of SNPs that correspond to the STR pattern of
the members of the group. While an STR mutation is a completely
different thing than a SNP mutation, the ancestors of the
members of the group that have the same STR mutations are the
same as the ancestors who had the associated SNP mutations. In
other words, the pattern of STR mutations is associated with the
chain of SNP mutations permits determination of a Haplogroup. A
Haplogroup is that group of men who have the same SNP in their
DNA, and goes by the name of the SNP. As you would expect, the
number of men in a early Haplogroup, say R-U106, is larger that
those in a descendant Haplogroup, say R-Z343. If you are a
member of R-Z343 you are also a member of R-U106, but the
inverse is true only for those few with the Z343 mutation I
looked up the estimate of when each of these occurred. The
earlest, U106, was about 4800 years ago and the latest, Z343,
about 2800 years ago. Don't think there were any surnames like
Douglas 2800 years ago.
If the entry in the Haplogroup column is in red, it means that
is FTDNA's conclusion as to what large Haplogroup the
participants belong based on their STR results. If the name of
the SNP in the Haplogroup column is in green, it means that the
participant has been tested and that is a real test result,
Within the Descendants of William group, there are three that
have tested SNPs, but two of them are very early SNPs (in fact
the two earliest in the chain of SNPs;that are in the header for
the group.) Only one has had a BigY test and therefore a TESTED
SNP that is more recent than over 4000 years ago. This is kit
N11063, who is Duncan Douglas. Duncan has two SNPs that occurred
later than Z343, namely BY13790 and Y62502. SNPs rate a name
when at least two subjects have them. So someone else besides
Duncan must also have these SNPs. We can try to find out who by
looking at a different project on FTDNA.
Besides surname groups, there are FTDNA groups based on
Haplogroup. We are interested in the one for R-U106. It is a
huge group with 21 pages of participants in the results. Here is
the link:
https://www.familytreedna.com/public/U106?iframe=ycolorized
.
Duncan is on Page 9, which you can get to by picking "9" from
the list of pages. Down toward the bottom of page 9 is a group
designated by a SNP chain "Z343>BY13790." There are three
Douglases and five others, two Parkes, one Rix, one Henricsson,
and one who didn't list a paternal line ancestor (surname) in
the group. You have to sign up for a Haplogroup project, and
don't have to have done BigY, and it looks line only three
Douglases did. Thankfully Duncan both did BigY and signed up for
the R-U106 project. We can observe that the last three in the
group all have a SNP downstream from BY13790, namely Y62502. The
Henricsson and Rix in the first group probably did BigY (because
they have a test result for BY13790) so really don't have
Y62502. I suspect the descendant of James Parke does have the
same downstream SNPs as the other Parke, he just didn't do BigY
and saved some money getting lesser tests .And I suspect that if
another descendant of our William 1610 goes for BigY that there
will be a number of downstream SNPs that match Duncan's and they
will receive names. And if that new tester joins the R-U106
project, he will show up next to Duncan in this table. The other
thing we can learn from this table is that the closest paternal
line relative to our William Douglass 1610 was named Parke and
was probably born in Ireland. He may be this guy
https://en.wikipedia.org/wiki/Thomas_Heazle_Parke
Other not-so-close relatives were from England and Scandinavia.
Nobody from Scotland.
I hope that is not too much of an expansion, I learned a lot.
Am I correct that once a random mutation in a DNA marker occurs,
all the male descendants thereafter carry the same mutated
marker?
As I said above, I have only encountered the word "marker" in
discourse about the chromosome positions where STRs occur. Male
descendants carry SNP mutations from the ancestor that had it
first, It is not a good assumption for STR mutations. Counts of
repeats at STR markers can go both directions. Julie recently
sent me a new Bubble chart where her Burt is shown with two
connections to the tree. I would have to study it some more, but
my opinion is that if the tree was based on SNP rather than STR
mutations, this couldn't happen.
Is a DNA marker mutation the same thing as a SNP? (I think I may
be in over my head! Help, please.)
If markers mean locations where STRs occur, Absolutely Not!
Locations where SNPs occur are chromosome positions and have
sequential numbers. Big numbers.
Didn't Bruce Linwood Douglass use these DNA marker mutations in
creating his bubble chart? It told him which of us had to
descend from a given ancestor and gave him an idea of between
which generations the mutation occurred.
Yes, and he did a great job. If STR counts only mutated one
direction, it should be completely accurate. With Burt having
two positions on the new chart, it must not be. And I have never
been comfortable with the fact that some of those tested had to
be sons of ancestors of William 1610. I am of the opinion that
all the participants in the FTDNA group named descendants of an
ancestor of William 1610 are really descendants of William. Part
of this opinion is based on the possibility of backward
mutations, but mainly on the fact that I don't think his English
surname was Douglas.
Is Duncan's Big Y test the only one to produce SNP results among
those who believe they descend from William Douglas (b. abt
1609)? What about Mark Brian Douglass (kit 113334, 37 markers,
SNP)? Malcolm Douglas (kit 156809, 111 markers, SNP, Gail
Goodcase, contact person)? On Julie's summary paper, "The 9th
Douglas/s Cousins, All Descended from William Douglas, b. abt
1610" updated 18 Nov 2012, there is a Robert Charles Douglas who
did a 111 marker DNA marker test in August 2017 but hasn't
responded to Julie. Did his 111 maker test produce SNPs?
Duncan's BigY test is the only one to produce SNP results. The
37, 67 and 111 marker tests produce that many STR counts. FTDNA
can estimate membership in a large, known Haplogroup from the
STR results, and they print it in Red in the Haplogroup column.
See answer to first question.
Would a few selected SNP tests among us William Douglas
descendants confirm what we think we know from paper trails?
Would it be worth the expense?
I suspect that Duncan's results have identified a number of SNPs
that are unique to him and now known only by a position number.
To be named, another probably Douglas would have to test, and
BigY is the obvious choice. Testing individual SNPs or packets
of SNPs just doesn't work so well. Better to go for it all to
start with. But be sensible in the choice of who. William 1610
had two sons, Robert and William. Whichever is Duncan's
ancestor, test a descendant of the other. I you are not sure
about the paper ancestry, test someone your are sure about. If
there are SNPs in common, you know that they occurred at the
conception of William 1610 or before and all descendants will
show these. Any SNPs that are not common occurred after the
conception of the son whose line they appear in. Testing of
individual SNPs can work if you want to prove some connection.
Doesn't work well if you want to discover a connection. This all
started with "I'll know it when I see it." That still applies.
Whether it is worth the expense is up to the Douglases
What about Bill Hough's statement that three of Duncan's SNP are
common with other individuals that don't carry the Douglas/s
surname? Significance?
Think this was covered in the answer to the first question. I
believe I was talking about BY13790 and was remembering the
conclusion from about 3 years ago. That count is now 5. If we
now talk about a downstream SNP, Y62502, there are two definite
and probably three, two Parkes and the anonymous guy. The
significance is that they are not unique to Douglas and occurred
before William 1610's birth, probably before surnames. More
significant for us would be SNPs that occur after 1610 which
will identify branches in William's tree of descendants. But it
takes a lot of testing and you don't know what to do until you
have results from at least one other descendant to compare with
Duncan's.
Be glad you are confined because of the virus, and have time to
absorb all this! If you have Douglas friends or relatives who
might be interested, please feel free to forward. either the
email or the attachment or both.
Bill Hough
|