| Author |
 Topic  |
|
FrankTompson
Starting member
United Kingdom
22 Posts |
 Posted - 15/10/2014 : 09:53:34
|
Hi Jan Peter,
Results received for Finn and I but am currently trying to make sense of them.
Will respond further after a bit more analysis,
Cheers,
Frank |
 |
|
|
FrankTompson
Starting member
United Kingdom
22 Posts |
Posted - 20/10/2014 : 13:33:51
|
Hi folks,
Herewith a screenshot of the basic results from FTDNA.
It effectively indicates that there is little/no relationship between Finn Stromme and Frank Tompson (yours truly). This is in contradiction to the DNADirectUK results (see page 43) which indicated a strong likelihood that we are related.
So, I continue to be confused. As I suggested earlier DNA testing just seems to be a free for all with no consistency in reporting methodology.
DNADirect showed a comparison of Alleles matches. FTDNA simply gives a %age likelihood of a match over 'X' number of generations, with no detail. I have asked Derek to go back and check if further detail is available.
In general Derek, Finn and myself have not found the FTDNA experience to be a particularly positive one, although Derek has managed to establish a working relationship with someone on the helpline (who seems to me to provide information that should have been readily available/accessible in the first place). (As an example, when I login to their site I can see 'results' (ie the screenshots posted above) for Finn, but not for me.)
All thoughts gratefully welcomed 
Cheers,
Frank |
|
|
|
JaneC
Norway Heritage Veteran
USA
3020 Posts |
Posted - 20/10/2014 : 17:09:44
|
Well first, condolences. So disappointing!!!
And kudos to you all for sailing into the brave new world of DNA testing.
I am curious what happens when the value of "1 generation" is changed to 5. Same results (probabilities)?
I am quite surprised, and find this to be the first time a paper trail was not confirmed to be correct. This is the most iffy paper trail I recall seeing, but still...
Couple of possibilities make it possible that Christian is Kristian, but the patience of the participants is likely exhausted and the chance of a refined, or revised, finding is too low to make pursuit of one an realistic option. For example, the possibility of a non-paternal event always exists. (Yes, Frank, you are adopted, LOL). Whether the paper trail for Finn was properly vetted, I don't know but have taken it on faith that it was.
About the comparison to the UK lab, we can see these findings confirm the relationship found. Frank and Finn share a high likelihood of a shared male ancestor within the last 12 generations or earlier. So indeed you are shown to be related.
The word related is used differently by different people in various contexts. As I think was earlier said (?), one prominent theory is that all humans share a common ancestor born in likely the thirteenth century - with the exception of limited, highly isolated aboriginal tribes. The sense of "related" we are going for is, related within five generations. So I take your statement, Frank, to mean the bottom line is that you and Finn are shown unlikely related at a distance of 4 generations (Frank) and five generations (Finn, presumably, although his relationship to this project has never been revealed to the forum. I just guess he'd be younger, further down the generations).
At FTDNA, if Frank and Finn had each tested blind, not knowing of each other, they would show up as a match to each other on a test of 37 markers and a genetic distance of 4, as shown. And correct me, people, if anything I say is a misunderstanding. We all know we're not experts here! So that is truly remarkable, in a sense. But the span of years to the point of relationship is too long for us to see where the point of connection may be. It seems that Daniel (father of Kristian) is possibly -but not likely - the point of connection.
Bummer!!
|
Edited by - JaneC on 20/10/2014 17:15:00 |
|
|
|
jkmarler
Norway Heritage Veteran
USA
7790 Posts |
Posted - 20/10/2014 : 17:15:00
|
Even though I was one who pushed for the additional test, I'm not certain that it helps much. It may be a man with more than one watch is never sure what time it is.
Per Jane's suggestion push the generation number to 4 or 5 to see if that changes the percentages.
|
Edited by - jkmarler on 20/10/2014 17:47:48 |
|
|
|
jwiborg
Norway Heritage Veteran
Norway
4961 Posts |
Posted - 20/10/2014 : 19:18:09
|
Hi Frank/Derek;
from the FTDNA results you received, do you have any info containing "Genetic distance" or similar, listed with a digit (0-37)...? |
Edited by - jwiborg on 20/10/2014 19:31:48 |
|
|
|
JaneC
Norway Heritage Veteran
USA
3020 Posts |
|
|
jwiborg
Norway Heritage Veteran
Norway
4961 Posts |
Posted - 20/10/2014 : 21:27:49
|
I missed the last sentence with the Genetic Distance info...
I don't think this was such a bad result at all!
A Genetic Distance of 4 means that there was a match on 33 out of the 37 DNA-marker tests (33/37 match).
How Is A Match Determined?
At 12 Markers
A 12-marker match is generally inadequate for genealogy purposes.
Most researchers will not respond to a contact about a 12-marker match.
Many researchers don't even allow comparison of their 12-marker results.
At 25 Markers
A 25-marker match should be at least 23/25 (Genetic Distance 2)
Many researchers consider this match to be sufficient if you share a surname.
At 37 Markers
A 37-marker match should be at least 33/37. (Genetic Distance 4)
We consider this to be the lowest matching level for those of different surnames (unless they have connecting paper trails.)
At 67 Markers
A 67-marker match should be at least 61/67. (Genetic Distance 6)
FTDNA recommends this test when comparing to others with a different surname.
This test is particularly useful in trying to better understand a large genetic family.
What Does a Match Mean?
For those who tested at 37 markers:
Less than 31/37 – the two participants do not share a common ancestor*
31/37 and 32/37 - the two participants have a small possibility that may share a common ancestor from the early days of surnames. This is an area with little clear insight. An upgrade to 67 markers is encouraged
33/37 - some researchers consider this to be a match and some don't. If there is a shared common ancestor - it will be more than a few 100s of years ago. Upgrade to 67 markers for additional clarity.
34/37, 35/37, 36/37 & 37/37 - the participants share a recent common ancestor |
Edited by - jwiborg on 20/10/2014 21:31:50 |
|
|
|
jwiborg
Norway Heritage Veteran
Norway
4961 Posts |
Posted - 20/10/2014 : 21:45:37
|
Derek/Frank,
with regards to your previous DNADirectUK test. That was a 12 marker test.
And as said, a 12-marker match is generally inadequate for genealogy purposes. 12 markers alone are usually not powerful enough to distinguish you from other family lines and can give inconclusive results. The more markers that you test, the more powerful your comparison becomes.
Scenario:
Mr. Jones has been studying his family's ancestry for several years and has started a "Jones" family study based in Arizona. He is interested in confirming that his family line is linked to a "Jones" line in New York. Although there are rumours that the two lines are related, Mr. Jones does not have the paperwork to prove this link. Furthermore, Mr. Jones has heard about the powerful capabilities of DNA testing for finding and linking family lines and would like to see if his line is linked to any other Jones lines worldwide.
Mr. Jones had previously chosen to test just 12 markers. After testing, he enters the 12 markers into a public database and finds out that he is a perfect match to the Jones line in New York. However, he also finds that he has a perfect match to over 200 individuals in the database, and over half of them do not even share his surname. How is this possible? Does it mean that he is related to everyone who matches him at the 12 markers? No, this simply means that data from only 12 markers are not powerful enough to distinguish Mr. Jones from other family lines.
To clarify this, Mr. Jones decides to upgrade his test to 20 markers. He enters the results of his 20 markers into the database and this time narrows down the number of matches, in fact, now, only 18 people match him perfectly at his 20 markers, including the Jones line in New York. Surprisingly, many of the individuals who used to match perfectly at 12 locations only match at 14 or less out of the 20 locations tested, confirming that there is no familial link with most of the 200 individuals identified in the 12 marker test (more than 3 mismatches indicates that two family lines are not related).
To further clarify the findings, Mr. Jones decides to upgrade to a 44 marker test. This time, he finds out that he is a perfect match at all 44 markers to only two lines, a Jones line in England, and a Jones line in the United States. After contacting the two lines and comparing paperwork and stories, Mr. Jones was able to confirm that his line was indeed definitely linked to both lines and he is now able to add both new lines to his family tree!
Surprisingly, Mr. Jones was also able to find out that only 43 out of the 44 markers matched with the Jones line in New York. This confirms that although the Jones line in New York is related to his line, they are more distantly related.
Mr. Jones also discovered that he had a close match to 4 other Jones lines (43 out of the 44 matched) and he is now pursuing the possibility that the 4 other lines are also distantly related to him (MRCA analysis dictates that 1 mutation occurs every 500 generations, and thus we would detect a mutation every 12 generations with the 44 marker test).
Mr. Jones is now trying to recruit more Jones males from throughout Europe to try to reconstruct and relink his family line.
Conclusions:
As you can see, the 12 marker test was simply not powerful and discriminating enough for Mr. Jones to pinpoint his family lines. After upgrading to the 20 marker test, Mr. Jones was able to obtain more useful information and was able to eliminate false matches generated by the less powerful 12 marker test. However, the 44 marker test was able to tell him the most useful information, and was able to pinpoint the people that he is looking for and was furthermore able to accurately answer his questions about his relationship to the Jones line in New York. Mr. Jones is now able to carry on his research, and as more and more people globally are tested and are added to the database, Mr. Jones will be able to reconstruct his family line in great detail and re-unite with Jones worldwide who are descendents of his family line. |
Edited by - jwiborg on 20/10/2014 21:47:53 |
|
|
|
jwiborg
Norway Heritage Veteran
Norway
4961 Posts |
Posted - 20/10/2014 : 22:01:09
|
A 33/37 match means "probably related".
You share the same surname (or a variant) with another male and you mismatch by four 'points' -- a 33/37 match.
Because of the volatility within some of the markers this is about the same as being 11/12 and it's most likely that you matched 23/25 or 24/25 on previous Y-DNA tests.
If you matched exactly on previous tests you probably have a mismatch at DYS 576, 570, CDYa or CDYb in our newest panel of markers.
If several or many generations have passed it is likely that these two lines are related through other family members.
That would require that each line had passed a mutation and one person would have experienced at least 2 mutations.
The only way to confirm is to test additional family lines and find where the mutations took place.
Only by testing additional family members can you find the person in between each of you...this 'in betweener' becomes essential for you to find, and without him the possibility of a match exists, but further evidence must be pursued.
If you test additional individuals you will most likely find that their DNA falls in-between the persons who are 4 apart demonstrating relatedness within this family cluster or haplotype. |
Edited by - jwiborg on 20/10/2014 22:01:31 |
|
|
|
jwiborg
Norway Heritage Veteran
Norway
4961 Posts |
Posted - 21/10/2014 : 00:27:57
|
Frank/Derek; do you have anything from your latest test (Y-37) showing any specific mutation info?
Looking at which markers where the mutation have occured, can give some additional information. If the mutation have occured on slow markers, there is a higher probability that the relationship goes far way back, and is not recent.
On the other hand; if the mutation have occured on fast markers (possibly marked with red), the relationship between the two testpersons are more recent, e.g a few generations. This is what you should look/hope for. |
|
|
|
JaneC
Norway Heritage Veteran
USA
3020 Posts |
Posted - 21/10/2014 : 01:43:42
|
Hi again, I agree with you Jan Peter that the finding as is - of a 33/37 match - is startling considering that a random guy in the UK and a random guy in Norway stepped up and got this match. It almost makes you think the iffy paper trail could be correct. If more markers were to be tested, and more individuals were to be tested, a still more complete picture would be rendered.
I think another way of putting the results goes like this: "If 100 men were tested and had these same marker values, more than a fourth of them (26.74%) will have a common ancestor who lived within the last eight generations."
And again I welcome corrections. But in any event, clearly there's a chance that Daniel Strømme could indeed be Frank's great-grandfather - but no guarantees. When you're in Las Vegas, at a gambling casino, what kind of gambler are you? I'm conservative with betting, but that's me. To continue on would mean finding additional descendants in Norway, strong-arming them into participating in this study, sending Derek and Frank the big fat bill for the additional DNA tests, and possibly reducing all the principals to gobbledygook-crazed shadows of their former selves.
That's one view.
OR it could be fun, it could be interesting and worthwhile, to explore further. It comes down to that personal choice. And who knows? It is indeed possible that by increasing the number of markers studied, the Frank-Finn results could turn out to be 107/111 - because the additional markers, if studied, might turn out to be a solid match.
Re. the fast markers vs. slow markers (relative to how quickly said markers typically mutate): I think the TiP tool that FTDNA provides has already factored that in. A genetic distance of 4 would generate better probabilities for one person and worse probabilities for another, depending on where those 4 mutations occurred (but perhaps Frank or Derek will address this). Frank's posted results say this:
"Since each marker has a different mutation rate, identical Genetic Distances will not necessarily yield the same probabilities. In other words, even though Mr. Frank Thompson has a Genetic Distance++ of 4 from Finn Strømme, someone else with the same Genetic Distance may have different probabilities, because the distance of 4 was prompted by mutations in different markers, with different mutation rates."
To sum up, I'm curious what would happen if the Christian Tromm DNA project continued on, but that's a lot to handle for Derek, Frank, and Finn, and becomes an even more personal decision than before.
P.S. I keep forgetting - Frank - about your complaint re the non-standard process etc. used by the various DNA labs, yes, that's just ridiculous!! |
Edited by - JaneC on 21/10/2014 02:04:07 |
|
|
|
jkmarler
Norway Heritage Veteran
USA
7790 Posts |
Posted - 21/10/2014 : 02:37:50
|
I see that some of the descriptions say more tests and / or more test subjects needed. In England, all the "between" generations, that is the children of Christian are passed on. In Norway, if I remember, Finn was a grandson of one of Christian's brothers. Also if memory serves, Finn is advanced in age, like 92 years. None of Christian's siblings live and so it's far from certain that there will be any more of the "between" generation to be found. There might be, not necessarily closer, other candidates down the line of Christian's uncle's line, and even further back for a connector. So the DNA test is not entirely nor exclusively as definitive as was hoped, for whatever reason.
The so-called "iffy" paper trail seems very clear to me. When I summarized the two sets of data (on pg 32, I think) that was before we had a paper record confirming Christian Thompson's birthdate of 2 Nov 1882 which is an exact match to Kristian's. Frank's father's note clearly shows Volden as the place of origin.
Since neither are 100% clear, maybe the two lean together enough to point out the correct conclusion, pending a better record or a perfect DNA match? |
|
|
|
JaneC
Norway Heritage Veteran
USA
3020 Posts |
Posted - 21/10/2014 : 02:59:13
|
Yeah, I definitely see what you mean about the paper trail.
As for more candidates, that would be another male line descendant in England and/or Norway, at the earliest possible generation but any generation will do. I don't know which would matter more, more people or an upgraded marker test.
Funny thing about statistical probabilities. You may have only an infinitesimally small chance of winning the lottery, but if you are the one who wins, you are 100% the winner.
I now see why Jan Peter asked about specifics re the specific markers. If Frank and/or Finn hit the lottery and some point in that line a mutation occurred on a slow marker, that would skew the probability to look "worse" than it should. Right? |
Edited by - JaneC on 21/10/2014 03:06:12 |
|
|
|
JaneC
Norway Heritage Veteran
USA
3020 Posts |
Posted - 21/10/2014 : 03:33:54
|
| Looking back at Jan Peter post, I see where that "inbetween" reference was made. I think Jan Peter is referring to a male of any generation whose Y-DNA profile falls between the outside parameters set by Frank and Finn profiles. If I am understanding correctly, this could be a later generation but presumably excluding offspring of Frank and Finn. |
Edited by - JaneC on 21/10/2014 03:49:14 |
|
|
|
NancyB
Medium member
USA
104 Posts |
Posted - 21/10/2014 : 04:58:51
|
I was alerted to this discussion by jkmarler. Some of you may remember that I posted previously about using FamilyTreeDNA to help an adopted friend, William, find biological family. (We think DNA rocks!) We used both the Y-DNA test (first 37 markers then upgrading to 67 markers) and the Family Finder autosomal DNA test. I think both are important.
William has a perfect match (Step 0) at 67 markers. But, that same person does not show up as even a distant relative on the autosomal Family Finder test. Both parties have also tested at 23andMe, and they don't show up as an autosomal match there either. We consulted with DNA expert Roberta Estes who confirmed what we were thinking: You can match 67/67 markers without having a close ancestor. The Y-DNA test is good for giving clues to a birth father's surname, but not necessarily for pinpointing relationships. That's where autosomal DNA shines.
In William's case, he is apparently only distantly related to his Step 0 Y-DNA match--not enough to even show up on Family Finder or at 23andMe. But, William is a confirmed first cousin once removed to a Step 1 37-marker Y-DNA match--confirmed through the autosomal test and then other resources. Another observation about Y-DNA results. In William's case, we noted a recurring surname among his Step 0, 1, 2 and 3 matches. However, this surname does not show up in his more distant Step 4 matches.
If the two parties involved in this discussion can upgrade to the Family Finder autosomal test, I would strongly encourage them to do so. If both have tested at FamilyTreeDNA, it only involves paying for the upgrade. Once those results are in, you can further expand the possibilities by downloading the raw data and uploading it to GedMatch.com. Not only will that compare the data with others who tested at various companies (who also upload their data there), but it may also have a lower threshold and include matches that didn't make the cut at FTDNA or 23andMe.
We found Richard Hill's Guide to DNA Testing Guide to be very helpful. It's a free download, and can be found here: http://www.dna-testing-adviser.com/
|
NancyB |
Edited by - NancyB on 21/10/2014 06:32:59 |
|
|
|
Topic |
|
|
|
Christain Tromm
