PDA

View Full Version : Genealogy Family Tree vs. Genetic Family Tree



Wing Genealogist
06-21-2017, 10:53 AM
A common complaint folks have in regards to the autosomal DNA test results from various companies is that the results do not match up well with their known ancestry. Almost all of the time, there is a direct or implied statement indicating something is "wrong" with the test itself.

While autosomal DNA testing is still not mature, and the creation of various "populations" still needs much more refinement, IMHO an even bigger issue is the fact we only inherit DNA from a very small portion of our ancestors even 400 years ago.

The nature of random recombination of DNA to create the sperm and egg means we do not inherit equally from all ancestors. We do inherit 50% of our DNA from our father and 50% of our DNA from our mother. But that is the only definitive statement we can make about DNA inheritance. We certainly do not inherit 25% of our DNA from each of our 4 grandparents. Each individual will inherit different, random, amounts of DNA from each of their 4 grandparents, and the percentages can vary widely.

Studies have shown where we have virtually a 100% chance of inheriting some DNA from all 16 of our Great-great-grandparents (Fifth Generation), but there is a 3.2% chance of not inheriting any DNA from one individual in the next generation (Sixth Generation) and a 35.84% chance of not inheriting any DNA from at least one member of your Seventh Generation. By the Eighth Generation, there is only a 49.5% chance of inheriting DNA from any one ancestor and it is theoretically impossible to receive some DNA from all 128 of your 5G-grandparents.

I am tracking which ancestors I have received DNA, and trying to look at how much DNA I inherit from each ancestor. I have cases where I have apparently received more DNA from a 5G-grandparent of the 8th Generation than I received from a 2G-grandparent from the 5th Generation. Another interesting fact: My mother descends seven different ways from a couple married in 1796. Despite these multiple lines of descent, I have not yet confirmed a single DNA match through this ancestral line.

The various autosomal DNA tests tell you they go back several hundred years to break down your ethnicity. Even if they only go back to 1500 (prior to the European colonization of the Western Hemisphere), they are only picking up DNA from an extremely tiny fraction of your ancestors (I believe much less than 1%). Given this, it is no wonder the ethnicities they report don't match up well with your genealogy paper trail.

Dave-V
06-21-2017, 12:37 PM
There are a couple of old threads where this subject came up and a few links to online articles on the topic were posted. But you're right that the odds of your genetic and genealogical ancestors being different rapidly increase after about the 6th generation.

That's one of the reasons several lines of my surname group haven't given up yet on their ancestral stories of being descended from "Indian princesses" even in the face of autosomal DNA evidence to the contrary... :P More interesting though, many of the US descendants of 18-19th century immigrant families are right at the 5-7th generation mark today or older and this issue can be a real problem for those trying to "cross the pond".

http://www.anthrogenica.com/showthread.php?6632-What-are-the-odds-of-inheriting-no-DNA-from-a-great-great-great-grandparent&p=145015#post145015

http://www.anthrogenica.com/showthread.php?4662-So-you%92re-related-to-Charlemagne-You-and-every-other-living-European

Pylsteen
06-21-2017, 01:38 PM
O yes, this is important to keep in mind, certainly when looking into lines to medieval nobility. I can have numerous lines to the same medieval noble, however, though he did contribute to my existence by giving birth to his child, he would have to compete with a few million other medieval people in being represented in my (less than a million?) genes.

sktibo
06-28-2017, 05:07 AM
A common complaint folks have in regards to the autosomal DNA test results from various companies is that the results do not match up well with their known ancestry. Almost all of the time, there is a direct or implied statement indicating something is "wrong" with the test itself.

While autosomal DNA testing is still not mature, and the creation of various "populations" still needs much more refinement, IMHO an even bigger issue is the fact we only inherit DNA from a very small portion of our ancestors even 400 years ago.

The nature of random recombination of DNA to create the sperm and egg means we do not inherit equally from all ancestors. We do inherit 50% of our DNA from our father and 50% of our DNA from our mother. But that is the only definitive statement we can make about DNA inheritance. We certainly do not inherit 25% of our DNA from each of our 4 grandparents. Each individual will inherit different, random, amounts of DNA from each of their 4 grandparents, and the percentages can vary widely.

Studies have shown where we have virtually a 100% chance of inheriting some DNA from all 16 of our Great-great-grandparents (Fifth Generation), but there is a 3.2% chance of not inheriting any DNA from one individual in the next generation (Sixth Generation) and a 35.84% chance of not inheriting any DNA from at least one member of your Seventh Generation. By the Eighth Generation, there is only a 49.5% chance of inheriting DNA from any one ancestor and it is theoretically impossible to receive some DNA from all 128 of your 5G-grandparents.

I am tracking which ancestors I have received DNA, and trying to look at how much DNA I inherit from each ancestor. I have cases where I have apparently received more DNA from a 5G-grandparent of the 8th Generation than I received from a 2G-grandparent from the 5th Generation. Another interesting fact: My mother descends seven different ways from a couple married in 1796. Despite these multiple lines of descent, I have not yet confirmed a single DNA match through this ancestral line.

The various autosomal DNA tests tell you they go back several hundred years to break down your ethnicity. Even if they only go back to 1500 (prior to the European colonization of the Western Hemisphere), they are only picking up DNA from an extremely tiny fraction of your ancestors (I believe much less than 1%). Given this, it is no wonder the ethnicities they report don't match up well with your genealogy paper trail.

How much progress have you made on this? How do you figure out which ancestors you've inherited DNA from, DNA relative matches?
Since we have about a 35% chance of inheriting no DNA from our 4th Great Grandparents, I'm thinking it might be a good idea to compare our DNA results to our pedigree only going back to 3rd Great Grandparents. I wonder if our DNA results would make a little more sense in this light? What do you think?

Wing Genealogist
06-28-2017, 11:04 AM
How much progress have you made on this? How do you figure out which ancestors you've inherited DNA from, DNA relative matches?
Since we have about a 35% chance of inheriting no DNA from our 4th Great Grandparents, I'm thinking it might be a good idea to compare our DNA results to our pedigree only going back to 3rd Great Grandparents. I wonder if our DNA results would make a little more sense in this light? What do you think?

Unfortunately, None of the current autosomal DNA tests on the market do a good job in allowing folks like myself to compare our "genetic" family tree with our "genealogical" family tree. Only Ancestry DNA really does a good job in providing you hints for DNA matches and lets you see the ancestors you share in common with these matches. However, these matches still need to be confirmed via two ways.

First of all, you should try to independently research the family tree of your match to make sure their "genealogical" family tree is accurate. I have had at least one case where I have a DNA match, and a family tree match, but when I researched the person's family tree, I found out the common ancestry almost certainly was not through the family tree match. My genealogical research showed a break in their lineage (where a supposed child was found to actually be a child of another couple). This was also confirmed by looking at the "Shared Matches" where all of the shared matches came through a different pair of ancestors in my "genealogical" family tree.

Even if you are able to independently verify the genealogical family tree is accurate, you then need to prove the DNA match actually came down through this match, rather than through another line. A chromosome browser would be a real help in this, so you would be able to verify whether your "genetic" matches to a presumed couple share roughly the same segments of DNA, or whether one (or more) of the genetic matches do not share any common segments with the other matches (which would indicate a high likelihood that your genealogical relationship to this match is actually through another line).

Even with a Chromosome Browser, you still would not be 100% certain a genetic DNA match is from one particular genealogical match.

For close relationships (roughly up to third cousins) the percentage of DNA you share with your match (along with your documented "genealogical" relationship) be all the proof you need (unless you happen to know you are closely related to your match via multiple lines).

Once you get out to roughly the fourth cousin level things become much more cloudy. By this time, the DNA inheritance pattern is so random you may by chance inherit more DNA from a more distant ancestor than you do from a closer ancestor. This would result in you sharing more DNA with a distant cousin than a closer cousin. In my research, I have found a 6th Cousin once removed whom I share more DNA with than a 3rd Cousin once removed. In this rather extreme level I believe it is likely I have another, currently undocumented relationship with the more distant cousin, but it is possible there is no other shared relationship.

What I have personally done is to create a modified Ahnentafel chart. Rather than listing each ancestor on a separate line, I list each couple on one line (as DNA really comes from couples, rather than individuals). I then looked at my Ancestry DNA matches and documented on the Ahnentafel chart when I am able to identify both a DNA match and a paper-trail match. I then have gone through the laborious process of creating multi-generational family trees of much of my family to independently verify the relationship (and along the way I have also been able to verify relationships to folks who either have no trees, or whose trees do not show our common ancestry).

Ancestry DNA states I have 435 pages of DNA matches, and counting them out this comes to 21,708 DNA matches at a predicted level of up to roughly 8th cousins. They also state I have 540 matches which are predicted to be 4th cousins or closer. Finally, they state I have 433 DNA matches where the online family trees indicate a paper-trail match. Some of these paper-trail matches have private trees, so I am not able to see the match. When this occurs, I contact the individual, let them know that Ancestry DNA has documented where we share DNA, and asked them to either invite me to view their tree, or at least to tell me our shared ancestors, our relationship, and their line of descent from our common ancestors (so I can independently verify the line)

To date, I have been able to document common "genealogical" ancestors for 509 of my DNA matches, but this still does not mean that our shared DNA comes down from the documented common ancestors. I have uploaded a copy of my DNA ahnentafel at:
https://app.box.com/s/64ft60edli7onov3eq2929k6i91i7abd

06-28-2017, 11:15 AM
I recently got a new DNA relative on 23andme, which implied 3-4 cousin with 1.09% shared DNA.
As this was my highest shared amount ever, I messaged them, and indeed their Grandmother was from Wales, with her name and maiden name, I did some research using Ancestry.co.uk and I found the link, turned out she was my third cousin once removed. A nice addition to my paper trail, and my first success with the dna relatives tool and finding also the paper trail.

Wing Genealogist
06-28-2017, 11:38 AM
I recently got a new DNA relative on 23andme, which implied 3-4 cousin with 1.09% shared DNA.
As this was my highest shared amount ever, I messaged them, and indeed their Grandmother was from Wales, with her name and maiden name, I did some research using Ancestry.co.uk and I found the link, turned out she was my third cousin once removed. A nice addition to my paper trail, and my first success with the dna relatives tool and finding also the paper trail.

Congrats on your first match!

firemonkey
06-28-2017, 04:30 PM
How much progress have you made on this? How do you figure out which ancestors you've inherited DNA from, DNA relative matches?
Since we have about a 35% chance of inheriting no DNA from our 4th Great Grandparents, I'm thinking it might be a good idea to compare our DNA results to our pedigree only going back to 3rd Great Grandparents. I wonder if our DNA results would make a little more sense in this light? What do you think?


If I go back to my 3gt ,as I have all of those, then Living dna has me pegged right at 100% British. With other tests I get a hotchpotch of different (minor?) populations.

firemonkey
06-28-2017, 10:15 PM
List of different minor/unusual populations I get


Southern Europe (23)
Ashkenazi jew (23)
Mediterranean islander (Dna land)
Finnish(Dna land)
Amazonian (Dna land)
Native American ambiguous(Dna land)

Dna tribes- Balto - North Slavic
Finnish
Lebanese Cypriot
Iberian
Balochi Punjab
Basque
Amazonian
Sephardic Jewish

My heritage- Finnish
South Asian
Native American

Gene plaza- East Meditarranean
North Slavic


Ancestry- Europe east
Italy/Greece


FTDNA- Finland
North Africa
North and central America
West Middle East

Dewsloth
06-28-2017, 11:06 PM
A common complaint folks have in regards to the autosomal DNA test results from various companies is that the results do not match up well with their known ancestry. Almost all of the time, there is a direct or implied statement indicating something is "wrong" with the test itself.

While autosomal DNA testing is still not mature, and the creation of various "populations" still needs much more refinement, IMHO an even bigger issue is the fact we only inherit DNA from a very small portion of our ancestors even 400 years ago.

The nature of random recombination of DNA to create the sperm and egg means we do not inherit equally from all ancestors. We do inherit 50% of our DNA from our father and 50% of our DNA from our mother. But that is the only definitive statement we can make about DNA inheritance.

^^Always good to keep in mind.
Just look at the apparent auDNA chaos M.O. 2.0 shows from just one generation. I "lost" all my Dad's British Isles (although some seems to have moved back to Scandinavia); and I'm not sure what's going on with the Sephardic.
In the abstract, for the last few centuries I should be roughly half Lebanese, 1/4 British Isles and 1/4 German, but:


Dad Mom Me
Middle Eastern 0% Middle Eastern 61% Middle Eastern 20%
Asia Minor
0% 31% 12%
East Middle East
0% 12% 2%
North Africa
0% 0% 0%
West Middle East
< 2% 18% 6%
European 94% European 21% European 52%
British Isles
47% 0% 0%:suspicious:
East Europe
0% 0% 0%
Finland
0% 0% 0%
Scandinavia
0% 0% 6%:noidea:
Southeast Europe
13% 21% 22%
Iberia
0% 0% 0%
West and Central Europe
34% 0% 24%
Jewish Diaspora 4% Jewish Diaspora 18% Jewish Diaspora 28%
Ashkenazi
4% 0% 0%
Sephardic
0% 18% 28%:confused: