Topics

spurious DNA match #dna


Arnold Chamove <arnoldchamove@...>
 

I recently found a very close autosomal DNA match with me on FamilyTreeDNA.
This person is a closer match to me than I am to my sister, and he is a
closer match to my sister than she is to me. We were born early 1940 in
San Francisco, California and he was born in Maine 45 years later. His
ancestry says 3% Jewish, 50% British, and 33% French/German; while my
sister and mine are reported as 95% Jewish (which is reasonable). My sister
and I share 2516 centimorgans (cM) while this man shares about 3384 cM and
similar with both me and my sister.
I assume such a close relationship can only if he and I share the
same parents--both parents. But we do not. Both my parents were dead when
he was born. This man is clear who his mother is. My sister and I look a
lot like our parents and not a lot like this man.
Is there explanation other than that the three of us share the same
parents?

Thanks
Arnold Chamove
Palmerston North, NZ


Steve <sjadelson@...>
 

Those kinds of shared numbers would indicate a parent/child
relationship with you, not siblingship. Or an identical twin,
which is clearly impossible. Otherwise, it looks like a lab error.

Steve Adelson

----- Original Message -----

From: "Arnold Chamove arnoldchamove@..."
Sent: Tuesday, June 21, 2016 7:41:35 PM

I recently found a very close autosomal DNA match with me on FamilyTreeDNA.
This person is a closer match to me than I am to my sister, and he is a
closer match to my sister than she is to me. We were born early 1940 in
San Francisco, California and he was born in Maine 45 years later. His
ancestry says 3% Jewish, 50% British, and 33% French/German; while my
sister and mine are reported as 95% Jewish (which is reasonable). My sister
and I share 2516 centimorgans (cM) while this man shares about 3384 cM and
similar with both me and my sister.
I assume such a close relationship can only if he and I share the
same parents--both parents. But we do not. Both my parents were dead when
he was born. This man is clear who his mother is. My sister and I look a
lot like our parents and not a lot like this man.
Is there explanation other than that the three of us share the same
parents?


Steve <sjadelson@...>
 

Followup on the "clearly impossible" identical twin: I should also
throw out the idea that this person is the contact for someone else
who actually took the Family Finder test, who is your long-lost twin.
Just to be thorough regarding the non-lab error possibilities.

Steve Adelson

---

Those kinds of shared numbers would indicate a parent/child
relationship with you, not siblingship. Or an identical twin, which
is clearly impossible. Otherwise, it looks like a lab error.


Arline and Sidney Sachs
 

Arnold Chamove wrote that his sibling and him both shared almost all of
their DNA with someone else and therefore they share the same both
parents. The is not the case. The amount of share DNA is computed as a
match if one of chromosome pair match one of the other person chromosome
pair. Therefore everyone is almost a perfect autosomal match to both
their parents, but the parents may have nothing in common. Therefore the
amount of DNA could come >from only one parent and not both.

The amount of share DNA >from autosomal testing for siblings is about 75%
which Arnold and his sister have. The unknown person have about 100%
matching autosomal DNA with both of them means that there is a
parent/child relationship with each and therefore could not be siblings.
Since there is a age differ, Arnold and his sister should be the parents
and the unknown person the child. The only way I can think of this is
possible is that the unknown person came >from Arnold sperm and his sister
egg. If the unknown person mother is alive, she should be tested.

About the ancestry estimates, the testing company used sampling of people
who had been living in the same places for a number of generations. The
results are their best guess without stating how long ago the estimate is
for. Since people did move around, the company may not have a perfect
data base. Therefore, it is not as reliable as the autosomal DNA results.

Sidney Sachs
Lorton, VA


Arnold Chamove <arnoldchamove@...>
 

I consulted an expert: the world-renown Steven Morse, who replied
telling me (and I paraphrase here):

The problem is in the autosomal testing. Unlike the yDNA or mtDNA
test which is an exact science, the autosomal results are based on
statistics. Although autosomal/Family Finder testing can sometimes
give good results, it can also give a totally false positive. It might
be that the matching segments are not >from any one ancestor but rather
from a combination of ancestors, and the mix of ancestors in Arnold's
DNA sample just happened to give the same DNA sequences as the mix of
DNA in Elliot's case.

So I wrote back asking (and I hope people here will comment also):

Here is how my brain has to sort this stuff out.
I view yDNA and mtDNA as follows: Millennia ago my ancestors wore a
set of distinctive clothing (my interpretation of DNA markers). All
their offspring wear the identical set of clothing. Every so often,
they replaced one item of clothing with a new item (mutation), and
pass that new outfit on to their descendants. Consequently, we
compare the number of items of clothing to analyse the relationship
between people. Easy-just count the number of identical clothing
items. [[There are two types of clothing in the male yDNA -- one
category of clothing that wears out fairly often, about every few
generations, like a cheap t-shirt (the STRs), and others that lasts
for at least thousands of years, like denim (the SNPs). The mtDNA, on
the other hand, has only the durable clothing that lasts for
millennium, i.e., slow-changing markers).]]

Autosomal: Here we only have 'samples' of the clothing items, some are
wool, some alpaca, some cotton; and some are dyed red and some dyed
yellow or green. Just >from these samples that we have for our
autosomal analysis, we try to calculate who is related to whom and how
close. But if someone has red cotton >from a shirt and another red
cotton >from trousers, they are spuriously regarded as somehow
identical, and this can be problematic.

I am pretty sure I have the yDNA stuff correct, but is my
interpretation of autosomal correct?"

Regards
Arnold Chamove
Palmerston North, NZ

---
I recently found a very close autosomal DNA match with me on
FamilyTreeDNA.
This person is a closer match to me than I am to my sister,
...