RootsChat.Com
General => Ancestral Family Tree DNA Testing => Topic started by: Gadget on Sunday 29 December 19 10:36 GMT (UK)
-
I manage my niece's DNA and have been contacted via My Heritage by someone who shares 69 cM over 3 segments. I thought I recalled the name so checked on Ancestry and find that the person is shown there as sharing 47 cM over 3. Exactly the same tree, name. location are used so I'm assuming this is the same person.
I know that different sites do differ slightly in the results produced but this seems a rather large difference and suggests that they are from two different but related people.
Any thoughts
Gadget
-
I forgot to say that the original test was via Ancestry. I then uploaded the results to My Heritage. Her results with me are:
Ancestry 1846 My Heritage - 1981
-
Gadget, what are the respective segment sizes? I'm sure you have considered this but I'm just curious.
Martin
-
I've found a few differences with my DNA on the two sites. I think My Heritage uses a different method of calculating shared DNA. These were also using the original Ancestry test results uploaded to My H.
e.gs of some of my relatives that I knew before I did any tests:
a.
My H 133 /7
Anc - 120/6
b.
My H 110/4
Anc 70/6
c.
My H 70/3
A 48/3
-
Gadget, what are the respective segment sizes? I'm sure you have considered this but I'm just curious.
Martin
I'm not sure that it makes much difference but this is what they are:
Shared segment sizes - 37; 16.6; 15.6
No way of knowing with Ancestry results.
-
Are the segments all a good (7cM+) size in each case?
Ships that pass...
Martin
-
I think you posted just after me - all are reasonable size.
-
I'm betting that all My Heritage matches have higher (and in some cases, much higher) shared DNA than Ancestry.
-
I don’t know if all do, but I can certainly confirm that at least a couple of mine have larger amounts of matching DNA on MyHeritage than on Ancestry. The one I can remember without checking is around 50CM on MH compared to around 19CM on Ancestry.
I have read that Ancestry seeks to eliminate the ‘background noise’ ie the kind of matching that occurs within a general population.
-
I have read that Ancestry seeks to eliminate the ‘background noise’ ie the kind of matching that occurs within a general population.
This would only be the case if the downloaded file from Ancestry included the 'noise' :-\
-
Speculating here; I don't know either way, but I think the raw data that you transfer from Ancestry to another site is your whole profile, not just the part that Ancestry decides to use in its calculations.
So thinking of your DNA test result as being like a book that Ancestry takes a set of quotes from to work with, and ignores the rest, what you send to MH is a copy of the whole book, which it can quote from as it chooses.
-
Like all the scatter points on a regression chart ;D
-
Ancestry's Timber algorithm probably accounts for the differences. Uploading to gedmatch is useful, eliminate hard breaks and look at the size and number of segments. If you have tested rather than uploaded to MH, gedmatch will show the differences between the sets of raw data. Just a suggestion for those more deeply interested...
-
That's fine but only if the match is on Gedmatch too.
-
Obviously. But for the purpose of illustrating the differences, I had, perhaps mistake!y, assumed that you might have tested at both ancestry and MH, or even elsewhere?
-
I have but many of those who match have not.
The most important thing is to link up to a common ancestor and that can often be done without a full inspection using a chromosome browser, etc. I find that such tools are often used to the detriment of good solid genealogical research and can often be used as displacement activity.
-
I don't disagree with your rationale. Not everyone wishes or needs chromosome details. The question posed appeared to be regarding the different cM/segments between matches one might have on ancestry compared to the same matches on MH. It's not clear how else they might be compared other than by using gedmatch. It just happens to be the case that an ancestry match will almost always show a greater cM if that match is also on gedmatch. You may be lucky enough to have access to more than your own raw data and try the above for yourself or pose the question elsewhere?
-
Thank you for your interest. I'm puzzled by your replies - especially suggesting that I
pose the question elsewhere?
which appears rather arrogant on your part.
I asked about the different results on Ancestry and My Heritage. The most logical answer was that Ancestry used post-processing methods to get rid of background 'noise' but My Heritage takes the data uploaded from Ancestry and uses a different technique or leaves the 'noise' as is. I'm quite happy to accept this.
In the cases discovered, we have identified the particular family line that the connection occurs in. I was just interested in why MH had a higher shared DNA than Ancestry.
Gadget
-
You will know that there are many sources of information provided freely by genealogists and others where themes/topics are raised/explored, questions asked/answered/expanded. Excellent as it is, Rootschat is but one.