Help with another DNA question, please!

Question

Help with another DNA question, please!

505 views

I got a strong match the other day, Both of us descend from a John, born in 1764. He had 2 sons:

William ,1802-1882
John, 1798-1844

William and John married sisters.

William married Esther, 1804-1854
John married Eleanor, 1800-1844

William and Esther's line (hers)

William 1802-1882, her 2x ggf
John , (1842 - 1882), ggf
Benjamin, (1875 - 1931), gf
Herbert (1908 - 1989), father
Cousin

John and Eleanor's line (mine)

John , (1798 - 1844) 4x ggf
Nicholas, (1840 - 1908), 3x ggf
Rebecca, (1870 - 1920), 2x ggm
Lena,(1897 - 1990), ggm
William,(1928 - 2008), gf
My mom
Me

Does this make us double cousins?? We share a total of 110.78 cM of DNA. That’s quite a bit for so many generations, right? Is this just completely bizarre? I'm stumped. Help! EDIT - ok, it appears we are 4th cousins 2xr. But this is just strange, seems like a lot of shared DNA for such a long distance between us.....

*Addition on 5/8/16 - Per GEDmatch guidelines: To qualify as a 'match' in the genealogical time frame, results must have a largest Autosomal segment that has at least 700 SNPs and be at least 7 cM.

Chromosome	Centimorgans Shared	Matching SNP
1	2.78	1300
1	2.06	500
1	26.12	5675
2	26.98	7200
2	5.45	1600
2	1.05	500
3	1.83	500
4	1.24	500
4	17.83	3500
6	3.69	563
8	1.58	500
10	2.88	700
10	3.49	700
11	1.5	700
13	1.77	500
13	1.43	500
13	1.94	500
18	7.14	2100

asked May 7, 2016 in Genealogy Help by Summer Orman G2G6 Mach 9 (94.6k points)
edited May 9, 2016 by Summer Orman

Summer, my dad, Herbert Fair Lunn's brother Harold Hunter Lunn, has a son, John Harold, my cousin, who took a DNA test. His wife Phyllis and I will get together online (they live in Clinton TN) and sort it out. She will contact me in the morning as she is not sure hat whether it is Autosomal or Y. If it is Y just think how much that would give us on our Lunns! John and Phyllis have one son and he has one son. This just keeps getting better thanks to you!

I'm thinking of getting an Mt-DNA test. What do you think? This would perhaps solve my brick wall. My great-great grandmother, Susan Sisco, born in 1818, had a child "out of wedlock" at 16 with my great-great grandfather, John S or T Turner who died in 1837 in Arkansas. Phebe was "given" to John's Aunt, Phoebe Ann Turner Sisco. Phebe is listed as a Ward of Phoebe Ann's husband, John Andrew Francisco. Susan remarried later in life and had several more children. There were many many Francisco/Siscos, and it may not ever be solved. An aside, the Franciscos (French Huegonots who fled to the Netherlands) go back to the settlement of Bergen County New Jersey, where my now husband's family lives!

commented May 9, 2016 by Jeanne Aloia G2G6 Mach 1 (19.1k points)

3 Answers

A Double Cousin will predict a closer relationship. So, in this case, a 3rd cousin prediction appears to be consistent with the paper trail.

What this tells me is that this match will further support the following relationships...

Hers...

John , (1842 - 1882), ggf
Benjamin, (1875 - 1931), gf
Herbert (1908 - 1989), father
Cousin

Yours...

Nicholas, (1840 - 1908), 3x ggf
Rebecca, (1870 - 1920), 2x ggm
Lena,(1897 - 1990), ggm
William,(1928 - 2008), gf
My mom
Me

What I mean but this is that given the existing paper trail, You inherited those segments from you mother, and she from Willliam, etc.

You don't know if a segment received by Nicholas came from John or Eleanor's line. Your cousin has a similar question.

commented May 7, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)

Answer 1 · 2016-05-08T00:22:20+0000

I'm not sure what you mean by "Wikitree excludes"

There is considerable debate in genetic genealogy over the smallest segment size that is valid and reliable for use in research. And, this is not simply a critique of the testing companies - genetic genealogists have done extensive studies of the validity and reliability of small segments (some argue that nothing under 5 cM should be used, others argue 7 cM, some have even disputed under 10 cM).

This is an important point for Summer because she is looking at a distant relationship, so the total amount of cM shared is a factor. Here is one of many articles out there critiquing the validity and reliability of small segments:

http://www.yourgeneticgenealogist.com/2014/12/the-folly-of-using-small-segments-as.html

commented May 8, 2016 by Ray Jones G2G6 Pilot (162k points)

Ray, I would use the same article to support my position. It's an article as to why you shouldn't manually calculate matches and genetic distance.

I would like to first point out that the cM sized used in the autosomal chart includes all segments, not just the larger ones. I believe it originally came from FTDNA. You are comparing apples and oranges. Wikitree does not add up the smaller segments, but the chart expects them.

The other point is that the DNA Services will give different weights to different segments. They don't use this chart, they compare segments against datasets for example or apply phasing which isn't apparent when you look at it from gedmatch. Gedmatch uses one national Data Set and FTDNA uses an internal Dataset. As these datasets improve and the algorithms improve, so will the predictions.

The article indicated small segments that the author matched with a cousin but her father did not. AncestryDNA would have clearly eliminated those unreliable segments by phasing and kept the ones that survived phasing. Even the larger segments may be eliminated or at least given less weight because they are believed to be IBS. Wikitree would keep them.

It is difficult for me to imagine that manually calculating genetic distance without all the other data and post-processing is better. When there is a conflict between the DNA services prediction and the Wikitree Prediction, I believe that the DNA services would be considered authoritative.

commented May 8, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)

Ken -

I'm sorry, but I completely disagree with you. It is not accurate to use small segments (particularly segments under 5 cM in length) as evidence in DNA genealogy. As Darlene explains below (and CeCe Moore explains in the article that I linked to), matching segments under 5 cM are typically IBS, and should not be used as evidence to support a relationship.

Yes - FTDNA reports all segments (including those under 5 cM) when you run its chromsome browser to compare two tests. But - Summer is going beyond a simple comparison and is trying to use DNA evidence to make a case in support of a second cousin - 2 X removed relationship.

I see this kind of thing all the time - typically in much more extreme cases in which a tester will try and "prove" a relationship from their family tree using a bunch of small segments. It simply is not accurate to use small segments as genealogical evidence, as segments under 5 cM are typically IBS.

Going back to Summer, if she uses the viable DNA segments (the 3 on chromosome 1, 2 and 4 - and possibly the 7.14 cM on 18), then there is a reasonably good chance that she is working with DNA segments that were passed down from her common ancestor with the match.

And - I agree with Darlene's comment below, in that if she focuses on these 3 or 4 viable segments, she may be able to find additional relatives who share one or more of these segments to achieve triangulation.

commented May 8, 2016 by Ray Jones G2G6 Pilot (162k points)
edited May 8, 2016 by Ray Jones

Ken -

You are completely confusing the science behind FTDNA's DNA matches and the separate matter of a genealogist having to use DNA evidence to support a specific family relationship.

Again - I am not disputing the fact that FTDNA includes small segments in its matching. They definitely use small segments in their matching.

For a genealogist (in this case Summer), she has to make a case using DNA evidence to support a claim. And - in her case, whether she is making a claim for 4th cousins - 2 X removed all the way down to second cousins - 2 X removed, she would be unwise to rely on DNA segments under 5 cM as evidence because there is a preponderance of evidence showing that most DNA segments under 5 cM are IBS.

You can't treat an FTDNA match as the same thing as an argument using DNA evidence to support a relationship. This is not the same thing.

It is simply not accurate to use segments under 5 cM to support an expected relationship between 5th-2nd cousins. The three large segments and the 7.14 cM segment are the only acceptable possibilities (with there even being reasonable doubt as to whether the 7.14 cM segment is actually IBD).

commented May 8, 2016 by Ray Jones G2G6 Pilot (162k points)

Ray, Thank you for your clear and direct responses. We agree to far more than you currently realize. Let me go step by step using your response.

"You are completely confusing the science behind FTDNA's DNA matches and the separate matter of a genealogist having to use DNA evidence to support a specific family relationship."

In this case, the specific family relationships in wikitree are to further support a single child to a single parent. In this case, she writes

"I'm just happy to have confirmed my mother's bio father from this."

She does not care about the specific relationship she has with her DNA match. And Wikitree guidelines indicate they don't care about that cousin relationship either.

"in her case, whether she is making a claim for 4th cousins - 2 X removed all the way down to second cousins - 2 X removed, she would be unwise to rely on DNA segments under 5 cM as evidence because there is a preponderance of evidence showing that most DNA segments under 5 cM are IBS."

She is not making this claim. Here is what she is claiming.

Her first claim is that the FTDNA DNA Service Prediction of a 3rd cousin,(I am sure there was a range), further supports the documented mother-daughter relationship between her and her mother.
Her next separate claim is that same FTDNA Prediction further supports the documented father-daughter relationship of her mother to her maternal grandfather.

She does not care about the specific relationship with the match, and no additional testing, as required by wikitree, is going to change these two conclusions. There is no reason to look at segments to support these two specific claims.

Are you saying there is a possibility that the 1 additional wikitree test will somehow refute these two claims? or that by looking at the segments, you are going to get a different result?

I agree, the next step would be to look at the segments to gather additional information. Genetic Genealogists would use this documentation and 1 DNA match to begin mapping the chromosomes. I have no problem in setting a minimum cM size of which segments get mapped.

I have never heard of a problem with mapping chromosomes up to but not including the Most Recent Common Ancestor(s). I have certainly never heard that the mapping of this part of the tree can not be relied upon because it lacked triangulation.

commented May 8, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)
edited May 9, 2016 by Ken Sargent

GEDMatch results are in
Minimum threshold size to be included in total = 700 SNPs
Mismatch-bunching Limit = 350 SNPs
Minimum segment cM to be included in total = 7.0 cM

Largest segment = 27.4 cM
Total of segments > 7 cM = 72.1 cM
Estimated number of generations to MRCA = 3.8 (not quite 4th cousins 2xr - I'd say this is a pretty good confirmation of Double 3rd cousins! Not that it matters anyway, we don't care!!)

644215 SNPs used for this comparison.

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
1	194,121,260	214,972,234	26.9	5,218

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
2	51,408,366	79,119,075	27.4	6,680

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
4	157,571,507	175,311,341	17.7	3,322

commented May 10, 2016 by Summer Orman G2G6 Mach 9 (94.6k points)

Summer, IMO your latest results merely restates what we already knew. What was the exact prediction on FTDNA, including the range? If you have already confirmed the relationship, you may have to look at the downloaded match file.

IMO, you have established your relationship with your match via a paper trail and confirmed that no NonParental Event has occurred on those profiles I listed above. Unfortunately, Wikitree does not agree that you have "Confirmed with DNA" any of them.

They are mistaken and here is why.

Wikitree users are confused as to the auDNA environment and how the Science behind auDNA matches relates to Wikitree.

There two primary reasons for analyzing matches. (1) The first is to establish a family relationship between matches without a paper trial, and (2) the second is to establish that a NonParental Event has not occurred for those with a paper trial.

Wikitree is applying the rules of (1) the prior to establishing the truth of the (2) latter, which is the wrong thing to do.

You have been referred to links of the importance of certain segments and how they should be used. All of these links are used to prove (1) a relationship between matches without a paper trial. Many of the same authors, many now are quite well known, have been with 23andme, like me, since auDNA was 1^st introduced. Their papers and postings for proving and not proving a relationship were in response to theories.

Postings and writings that involved matches with known documented relatives were almost always on the subject of mapping chromosomes, not proving match relationships. I have never seen anyone write that you needed to prove a relationship with DNA before you could then map chromosomes.

An example of this demonstrates my Point. http://www.yourgeneticgenealogist.com/2014/12/the-folly-of-using-small-segments-as.html

In this Article, the argument is that...

“people tend to see evidence that supports their theories and reject evidence that does not.”

These people are those who do not have the documentation to support their claim. These are theories. This article is directed at them, not those who have a documented relationship.

Unless Wikitree can accept and admit they were confused by this, it won’t be of any value to explain the rest.

Edit: Just in case the connection to wikitree was not apparent. There is an estimated 1 to 5% possibility of a Non-Parent Event occurring in any well-documented tree. When there is a match that supports the tree, that percentage drops. I am estimating to ~0%. Genetic Genealogists map the chromosomes for those connected in a match, up to but no including the Most Recent Common Ancestor(s).

For these same profiles, wikitree does not "Confirm with DNA"

commented May 10, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)
edited May 10, 2016 by Ken Sargent

Hi Ken,

I thank you for being so supportive on your post. I do have to laugh though, because there is so much discussion on errors or the possibility of Non Paternal events....when in fact, my mother WAS a non-paternal event. She didn't know William was her father until she was 30. He had known about her but while his wife was alive he couldn't admit he had strayed (though I'm told his wife did know). After his wife passed, he was able to welcome my mother and our family fully. I was standing at his bedside when he died.

So, as you previously stated, this relationship "confirmation" between Jeanne and I, serves 2 purposes: 1) To confirm (yes I say confirm) that my mother was Bill's biological daughter, and 2) I have a new cousin! We may be quite distant and only share a few DNA segments, but that doesn't mean a thing. We plan to have a relationship.

Summer

commented May 10, 2016 by Summer Orman G2G6 Mach 9 (94.6k points)

I just have to say, it is quite interesting that on the other cousin I posted about, the system confirmed us as 2nd cousins twice removed (consistent with the paper trail) but with Jeanne, the paper trail overrides the DNA Match.

James and I share over 120 cM but only 103 is counted in the total because the rest comes from smaller segments.
Jeanne and I share over 100 cM total, but only the segments totalling over 84.5 are counted.
there is virtually NO difference in the MRCA estimation, the only difference is James and I have larger segments in common, whereas Jeanne and I have larger segments overall, but more small ones in common. Yet, our relationship isn't recognized, even with the paper trail as proof.

Interesting.

	Total cM shared	largest cM shared	generations to MRCA
Me-James	103.5	26.2	3.6
Me-Jeanne	84.5	27.5	3.7

commented May 10, 2016 by Summer Orman G2G6 Mach 9 (94.6k points)
edited May 10, 2016 by Summer Orman

Summer, You may find this interesting...

The ability to confirm with DNA is all or nothing.

Parent-child relationships can be labeled Confirmed with DNA when there is a sufficient amount of shared autosomal DNA between two relatives who are third cousins and closer. It is necessary to use triangulated groups — three or more matches on the same DNA segment — when the testers are greater than third cousins.

A sufficient amount of auDNA is determined by the total number of cMs of several segments normally shared specific relatives. See Mathematical Average for auDNA Sharing andAre There Any Absolutes in Genetic Genealogy?" see help.

You only care about the amount of shared autosomal DNA between two relatives who are third cousins or closer. You and Jeanne are not third cousins or closer. The guidelines for you and Jeanne are...

"If all the test-takers have uploaded to GEDmatch, use their Triangulation utility or One-to-Many utility (for you and each segment sharing cousin). Find segments measuring 7 cM or more that you share (i.e. overlap) with two or more distant cousins.

Those who all match each other on the same segment form a triangulated group (TG). Using WikiTree's Relationship Finder find the shared ancestral couple for that TG." see help

There are a few interesting things to point out. The requirements only care about triangulated groups which 3 distant cousins share a single 7cM or greater segment. You don't need to predict the relationship between you and Jeanne.

1. This means that you only have to find one additional distant cousin who shares a single 7cM triangulated segment, to mark all the profiles up to and including the 2 shared ancestral couple. This is ok when you don't have a paper trail and are looking for clues as to how you are related, but it certainly falls short of proving an existing relationship.

2. You may be a 3rd cousin 1x removed but these rules allow for you to mark all profiles, even if the prediction is distant, based on a single 7cM triangulated.

3. There is no requirement that the additional match needs to be more distant, which means that you are proving pretty much nothing.

4. The x chromosome is obviously wrong, but even if it were corrected. You could Confirm with DNA any match, including a 3rd cousin 1x removed, with a 1cM match on the X Chromosome, even if the prediction is NO RELATION.

commented May 10, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)

Peter, your statement "The Family Finder system only predicts or estimates the relationship based on the number and size of the shared auDNA segments." is incorrect.

In March of 2013, see blog

the latest version of the autosomal DNA reference data set being used - for the past two months in order to refine their matching algorithms and provide better matches to their customers. Many customers saw the following message from Family Tree DNA on their homepage this month:

You may notice that some of the matches you had previously are no longer listed. In almost every case, those matches are in the distant and speculative range. Some may be legitimately related, but many were likely “noise”—small coincidental matching segments that create the illusion of matching. The more distant the relationship, the less likely a match can be predicted with confidence. We are aware that you may have spent a significant amount of time researching those matches, but Family Finder was specifically designed to find matches within the past 5-6 generations.

As the field continues to advance, we're obligated to continually apply the latest scientific advances not only to current and future tests, but to those that have already been performed. Sometimes this will involve changes to your results. Because this particular field is still evolving, it is likely that future adjustments will be made to refine your results.

Gedmatch. see FAQ

Autosomal Comparisons

Q. Why does the GEDmatch.Com Autosomal comparison give slightly different results than I get from my testing company?

A. FTDNA uses proprietary 'sets' and 'cM' tables when generating their comparison results. FTDNA's method results in segment endpoints tied to the endpoints of their 'set' data. The GEDmatch.Com algorithm is not dependent on 'sets', and will generate segment endpoints immediately adjacent to points where individual alleles don't match. Genetic Distances (cM) on GEDmatch.Com are calculated based on charts produced by the Computational Genetics group at Rutgers University.

23andMe and other testing companies also use their own comparison algorithms, and Genetic Distance charts. So, results are bound to differ than what you see here and on other 3rd party sites.

Why is only looking at total cM and then comparing to a generic chart, better?

commented May 10, 2016 by Ken Sargent G2G6 Mach 6 (62.0k points)

Answer 2 · 2016-05-08T15:50:18+0000

Hi Summer, I agree with Ray as far as the three substantial segments. The 7.14 cM on chromosome 18 is very possibly IBD, though. The rest are most likely IBS (identical by state).

You can't be sure of the common ancestor that you're getting this shared DNA from until you have a triangulated group, whereas someone else that descends from John (born c 1764) and his wife matches both you and your cousin on one (or more) of your large shared segments, i.e.the three large segments on chromosomes 1, 2 and 4.

My mom shares 73.7 cM with one cousin, and 126.7 cM with another cousin; they all three share common 2x great-grandparents. Another cousin shares 98.7 cM with my mom, and they share common 3rd great-grandparents. So you can see that the amount can be high even with 3rd ggps (great-grandparents).

I'd say the bottom line is that you need to both get on Gedmatch and find/wait for triangulated groups to analyze and try to determine your MRCAs.

Categories

Help with another DNA question, please!

Please log in or register to add a comment.

Please log in or register to answer this question.

3 Answers

Please log in or register to add a comment.

Please log in or register to add a comment.

Please log in or register to add a comment.

Related questions