RootsChat.Com
General => Ancestral Family Tree DNA Testing => Topic started by: lydiaann on Saturday 24 February 18 16:23 GMT (UK)
-
Did anyone listen to 'You and Yours' on Thursday this week (22 Feb, 2018)? One woman was interviewed as she had written to say that different labs produce different results for the same DNA. You can hear the interview on-line on the BBC website but, in essence, her complaint was that she was tested at 2 different labs and got very different results from each. Maybe it's useful so you can link up with people with the same familial strand, but as far as 'ethnic identification' is concerned, it seems to be rather iffy.
-
I agree with everything you have written there, and indeed it echoes what many people have written on here in the past. Alas there are still those who believe that DNA results and in particular the 'ethnic identification' are infallible and scientifically 'proven'. oh well .............
-
I think it depends on what we are comparing. There are four things we might compare:
1. The actual DNA results. These will vary a little because different companies test different parts of the genome and different amounts of the genome, but basically the results will be very similar. We don't usually get to see this data, it is a long file of As, Cs, Gs and Ts which is pretty meaningless to us.
2. How much DNA we have in common with other people who have tested. There may be slightly larger differences here between companies, partly because of the different amount of DNA tested, but also because some companies make up for the fact that they aren't sampling at as many points by (if I understand it correctly) "imputing" data that isn't actually tested, but which is statistically likely. I tested with FTDNA and Ancestry and my results here were very similar, only minor differences.
3. Then there is the prediction of the likely relationship between us and our matches. This varies even more between companies because of the differences in #1 and #2, but also because the different companies use different methods to estimate relationships. There is no "right" and "wrong" here because this is still an evolving science, though it would be interesting to see data on how each company predicts relationships that are known. Overall, I have found FTDNA to get this pretty right in a number of cases I can check, and Ancestry did OK too in the few cases there that I could check.
4. Finally, there are the ethnicity estimates. This is where the variation between companies is likely to be greatest because, again, they all use slightly different algorithms to calculate ethnicity. Again, there is no "right" answer, especially as our ancestors may have come from UK 5 generations ago, but Finland 15 generations ago and Ukraine 100 generations ago (as seems to have been the case with my mtDNA). And of course back 5-10 generations, we might have hundreds of ancestors, and they might come from all over the place, making assessment even harder. Everyone (except maybe Ancestry's TV ads) will tell you ethnicity is reasonably accurate at the continent level, but still not very reliable at the country or regional level.
So we should expect #1 and #2 to be pretty accurate, #3 to vary a little and #4 to vary a lot.
-
Did anyone listen to 'You and Yours' on Thursday this week (22 Feb, 2018)? One woman was interviewed as she had written to say that different labs produce different results for the same DNA. You can hear the interview on-line on the BBC website but, in essence, her complaint was that she was tested at 2 different labs and got very different results from each. Maybe it's useful so you can link up with people with the same familial strand, but as far as 'ethnic identification' is concerned, it seems to be rather iffy.
Yes, I heard this and was very interested. It confirmed my thoughts on DNA testing:
http://www.bbc.co.uk/programmes/b09rwt8d
Romilly.
-
Hi
One thing - DNA Testing is not a substitute for family history research ...
http://www.rootschat.com/forum/index.php?topic=786838.msg6443596#msg6443596
-
Folks are all caught up in Ethnicity when its really of no benefit to your research. There is a lot of negativity surrounding this and has more years of advancement to happen before consistency, better accuracy. Now even if it was accurate for everyone, it still offers, for the most part, very little benefit.
When you get your DNA results they are split into 2 parts - Ethnicity and Matching, its the matching that is useful for your family history research.
Eric has explained this in more detail above.
Hi
One thing - DNA Testing is not a substitute for family history research ...
http://www.rootschat.com/forum/index.php?topic=786838.msg6443596#msg6443596 (http://www.rootschat.com/forum/index.php?topic=786838.msg6443596#msg6443596)
Its another tool that works in conjunction with your paper research.
-
Folks are all caught up in Ethnicity when its really of no benefit to your research.
...
Hi
One thing - DNA Testing is not a substitute for family history research ...
http://www.rootschat.com/forum/index.php?topic=786838.msg6443596#msg6443596 (http://www.rootschat.com/forum/index.php?topic=786838.msg6443596#msg6443596)
Its another tool that works in conjunction with your paper research.
Precisely.
http://www.rootschat.com/forum/index.php?topic=758922.msg6100988#msg6100988
"the quality here will be dependant on what effort you and the other people put in on there trees to help find the connections,"
----------
I was looking at the Tree and Arms of the female and family, who my family married into, in December 1851.
Their Family have made a panel with Arms, as a photograph is online.
I noted one of the locations of their family was a HALL name, so I looked it up in 19th Century Books and the family who lived at that HALL, wasn't theirs!
According to the England & Wales Census they lived at the ... Ho. / House, of the same name in the same village.
-----------
There are even a few Rootschatters on these threads admitting to online tree errors, which they have not yet got around to correcting.
----------
DNA Testing might only lead to more confusion, if one of the parties, has not worked backward properly and methodically with actual Official Copy records, actual Registers, Census, Wills and other records, from their earliest known family in their current country.
Mark
-
-----------
There are even a few Rootschatters on these threads admitting to online tree errors, which they have not yet got around to correcting.
----------
Its a brave (or foolish) rootschatter who states categorically that their tree contains no errors, errors creep in despite our best intentions.
And yes I have an online tree with errors in it, but it is private to stop people copying the errors, although I wonder why when I see trees that have children born to mothers who died several years before the childs birth! (I know where the errors in my tree come from and that is when my family tree software got corrupted. Arghhhhh !)
-
... although I wonder why when I see trees that have children born to mothers who died several years before the childs birth! ...
This is the kind of point I was trying to make.
Mark
-
DNA Testing might only lead to more confusion
But I think more often it leads to clarity. Besides adoptions with unknown parents, which can really only be solved through DNA except in exceptional circumstances, there are many cases (some say somewhere between 10% and 30% of births) where the named father (generally the husband of the mother) isn't in fact the biological father. I know of a case where it seems likely (not certain, as DNA hasn't led to an answer yet) that the named father was paid to accept paternity. Probably all of us with trees that go back more than 2 or 3 generations have false paternity "errors" in our trees but in many cases we don't know, and again, DNA or some written "confession" are the only ways to find out.
Surely this is an unnecessary argument. Both paper and DNA genealogy are important, both are necessary or very helpful in most trees. Being aware of the dangers and pitfalls of both is obviously important.
-
Excluding Ancestry, which has least helpful DNA tools, the others (geni, myhertiage, gedmatch, livingDNA (this year) are rapidly improving the helpfulness of the tools available to assist us with connecting to other.
For example, another update for Myheritage - https://blog.myheritage.com/2018/03/major-upgrade-to-the-myheritage-chromosome-browser-for-better-exploration-of-dna-matches/
-
I've just been playing with the MyHeritage chromosome browser Bobby. It's certainly an improvement on what they had, and I was thrilled when they added a chromosome browser a few weeks ago. It almost makes up for a 2nd cousin whose DNA is incredibly useful who isn't keen on transferring from MyHeritage.
Hopefully in the future there will be a way to just select shared matches and put them straight in the chromosome browser.
I had been viewing each match on particular lines one by one and then uploading to spreadsheets. This will involve far fewer downloads and cut and pastes.
-
Those interested in My Heritage may be interested in the rather exciting announcements in this post on DNA Explained (https://dna-explained.com/2018/03/02/day-2-rootstech-vendors-visits-and-myheritage-is-smokin-hot/) - down towards the bottom.
-
Those interested in My Heritage may be interested in the rather exciting announcements in this post on DNA Explained (https://dna-explained.com/2018/03/02/day-2-rootstech-vendors-visits-and-myheritage-is-smokin-hot/) - down towards the bottom.
Thanks for posting the link. I had a read and particularly liked the DNApainter. I watched the YouTube presentation by one reviewer and it ties in with an article I once saw on the blog run by rootschatter "DevonCruwys" https://cruwys.blogspot.co.uk/ where one person had constructed a similar painted effect the old fashioned way using spreadsheets etc. I see the cruwys blogspot also has an interview with Jonny Perl who developed the DNAPainter tool
(if I understood the youtube presentation I mentioned then DNAPainter is easier to use with results from companies like 23andme, ftDNA but more difficult to use with results from Ancestry (you have to request additional information from ancestry before you can use it)
I do not like the way these DNA Genealogy companies are creating speculative trees because although they are based on DNA results (OK) they are also including names dates and places culled from peoples trees and as we all know there are a lot of mistakes in some trees
-
A similar fantastic(!) idea discussed here:
http://www.rootschat.com/forum/index.php?topic=788974.0
Carol
-
For those that would like to try DNApainter https://dnapainter.com/about
It greatly benefits your research... just keep adding those chromosome matches.
-
DNA Testing? Oh Yes - Oh No!!
So far my Niece (my Sister's Daughter) has only managed to take our tree forward!
She contacted me and said she had a Spencer match, I was quaking with absolute excitement!
I had two Spencer Marriages in the same line to a Hood, in the early to mid 18th Century (which I just can't link back to), but also Spencer on my Mother's side in the 20th Century.
I'm afraid the DNA relative was a descendant from the 20th Century Spencer, but obviously very pleased to hear from her!
My excitement evaporated when I realised I knew her late Grandfather very well (he used to come and stay with his Sister, my late Gt Aunt and they came to our house for meals). But glad to see we both have the same trees (with Certs etc.) going back 200 years, on my Mother's side.
So that is my brush with DNA!
Mark