Re: Blog on Z17911 STR Tree

James Bennett

Fantastic, thanks!

Sent from my iPad

On Jan 25, 2017, at 10:46 PM, Jared Smith <jared@...> wrote:

James -

Here's a long writeup that may be helpful.

STR mutations are passed down from father to son, with a change in an
STR value happening every few generations (more or less). Analyzing
STRs can tell you generally how related two people are. The more STR
values that are the same or very similar, the more closely you are
likely to be related. The more STRs you test, the more accuracy you
get. We've all tested to at least 67 STRs, so we have pretty good
accuracy. But STRs fluctuate a bit and change somewhat regularly (some
more than others), so it's difficult to place an exact timeline or
degree of relationship to the individual mutations.

What Joel is analyzing is the differences between the STRs each of us
have and the most common STR signature (mode values) we all have (or
groups of us have) collectively. The more different someone's STRs are
from the mode values of everyone else in a group, the more distantly
related they likely are. That's what Joel's charts are showing - an
estimation of relationship over time based on STR similarities and

STR comparisons are what have allowed us to find people that probably
have the R-Z16357 SNP mutation because their tested STRs are very
similar to others that we know are R-Z16357.

SNP mutations generally occur with one specific man and are then
passed down forever to his grandsons. As these distinct mutations are
identified, we can build a tree that shows a history of mutations
(typically one mutation occurs every 100 years or so). Each distinct
mutation is assigned a haplogroup name (e.g., "R-Z16357").

Big-Y testing identifies lots of SNPs - some shared with others and
some that only that Big-Y tester has (called 'private' SNPs).

The difficulty with SNPs is that it takes at least two people getting
positive tests for a particular SNP to verify that it came from a
common ancestor - and to start to assign a timeline to when that
mutation occurred. If an SNP isn't shared among relatives (it's a
private SNP), then it doesn't tell us anything particularly useful (at
least initially until perhaps someone else tests positive for it then
it is no longer private to only one person).

Merrick was identified as having the BY11565 private SNP - nobody else
had yet tested positive for it. Goff's test results just show that he
has the same BY11565 SNP. So this verifies that it came from a common
ancestor and isn't a very recent, private SNP at all. Because BY11565
is shared by only Goff and Merrick, but BY11573 is shared by Bennett,
Thomas, Goff, and Merrick, we thus know that BY11565 is more recent
than BY11573 - a Goff/Merrick ancestor formed this mutation, but not a
Bennett/Thomas/Goff/Merrick ancestor.

Because BY11565 is the most unique, yet also shared SNP they have,
this is their new terminal SNP (the word "terminal" being used a bit
loosely because it can change).

As another example, Joel Hartley has a current terminal SNP of Z17911.
He has also tested positive for the (currently) private A11130 SNP
that nobody else has yet tested positive for. So we know A11130 is
younger than Z17911, otherwise others of us would have it too. We just
don't know how much younger it is.

To try to 'prove' an SNP as being from a common ancestor, you want to
test the most distantly related person you can find that shares that
known or potential male-line ancestor. If Joel's brother tests
A11130+, this would tell us nothing new because A11130 may have
occurred at their father. If his 7th cousin tests A11130+, then we
would know the SNP is at least as old as their 6th great-grandfather
(and younger than Z17911).

So, the goal of our project is to identify distantly related cousins
(based on surnames, genealogies, and/or STR relationships) to try to
both identify (via Big-Y) and 'prove' (via Big-Y or other SNP tests)
younger and younger SNPs/branches. Each new shared SNP/branch moves us
closer to present day.

Eventually our tree will have a whole suite of shared SNPs identified.
If someone is an STR and/or surname match to someone else in our
group, an inexpensive SNP test could then establish exactly how they
connect to our tree on that line, and thus the rest of humanity. This
is the end goal - and we're making good progress toward getting there.


On Tue, Jan 24, 2017 at 8:42 PM, James Bennett <james@...> wrote:
Interesting! But over my head for most of it. Any good links to STR and SNP for Dummies?

Tnx, james

Sent from my iPad

Join to automatically receive all group messages.