RootsChat.Com
General => Ancestral Family Tree DNA Testing => Topic started by: ikas on Wednesday 16 August 23 09:31 BST (UK)
-
Ancestry originally assigned two of my matches to both sides. I last checked about two months ago and it was still only two matches. Checked this morning and they have now assigned 78 to both sides. Has anyone else noticed a sharp increase in assignment to both sides? If so, do you think the asssignments are largely correct? I have reservations about mine. Gedmatch suggests my parents were not related. Of the 78 now assigned to both sides, some I know are on my paternal side and some are on my maternal side. Seems unlikely to me.
-
They appear to have done an August update and I now have an additional 5 matches assigned to both sides - none of which I know.
However 78 seems a high number to assign to both sides.
Pheno
-
Parental assignment uses a computer algorithm involving known DNA matches and a trawl of Ancestry family trees to recognise lines of descent and match them between trees and hence to families. It may be subject to the same errors as Thrulines where user trees are incorrect and I think it is still in beta. It is not derived purely from an analysis of your DNA. Autosomal tests are not phased and so the phasing is estimated.
Ancestry do say that it will be at least 90% correct for around 9 out of 10 users, but less than 90% correct for the other 1 in 10.
There is always the possibility that a new update to the algorithm has introduced an error...
Edit: Just looked and my allocation for both sides has gone from 1 (my brother) to 15. The other 14 are all below 10 cM matches and I suspect an IBC error as a result of their updated algorithm.
-
Thanks for your replies. So, likely the update in August was the cause.
Phil57. Surprised by your comment that assignments are made based on trees as well as segments. The original statement said it was wholly dependent on shared segments. Do you know if they have changed the way they make these assignments?
-
Ancestry originally assigned two of my matches to both sides. I last checked about two months ago and it was still only two matches. Checked this morning and they have now assigned 78 to both sides. Has anyone else noticed a sharp increase in assignment to both sides? If so, do you think the asssignments are largely correct? I have reservations about mine. Gedmatch suggests my parents were not related. Of the 78 now assigned to both sides, some I know are on my paternal side and some are on my maternal side. Seems unlikely to me.
Gedmatch also suggests my parents are not related, which according to my paper trail I would agree with. They came from different parts of the UK. However they do not need to be related for you to have a "Both Sides" match. I have one which I found myself. Ancestry only assigns it to paternal side (the closest side). The person is my third cousin on paternal side. And 5C1R on maternal side.
His paternal grandmother was the granddaughter of my 2 x great-grandparents on my father's side.
His paternal grandfather was the 2 x great grandson of my 5 x great grandparents on my mother's side.
So if this match's father had done a DNA test, Gedmatch would also say his parents were not related to each other.
-
I had 1 "both sides" match that Ancestry chose before, today I have 11, the highest match is 15cm.
It has to be done on DNA and not trees, because only 5 of those 11 have a linked tree.
-
Phil57. Surprised by your comment that assignments are made based on trees as well as segments. The original statement said it was wholly dependent on shared segments. Do you know if they have changed the way they make these assignments?
Sorry, yes I've just checked and they are now saying that the assignments are being made solely from DNA, although originally they were utilising an algorithm to refine parental sides via trees. Probably, as they've increased the database and refined their processes, they've removed that aspect.
But they can't phase your DNA. Most autosomal tests including Ancestry's are unphased, which is partly the reason that lower length matches are less accurate. They are calculating the phasing of base pairs by comparison between matches who share the same string lengths. As the size of the database and number of comparisons increases, the accuracy should also increase, but the ultimate results are still governed by how they calculate and produce them.
https://support.ancestry.co.uk/s/article/SideView-Technology
I still suspect an issue with the latest update. As I say, my both sides allocations are all at low match lengths (apart from my brother) where the probability of IBC matches is relatively high.
-
I think there is something odd about this update.
My full sibling sister has an additional 27 'both side' matches but none of these are the ones that show up for me. I know that we won't necessarily have the same DNA but for none of these 'both side' matches not to be the same seems odd.
Pheno
-
I too have some both side.
The are 16 DNA matches who are in this class, highest is a mere 21cM with Zero shared match’s.
My Paternal and Maternal lines are from very different parts of England so the probability of this happening in my case in recent times is in any case very, very low.
Using the Gedmatch “Are My Parents Related” tool produces a NO.
-
I too have some both side.
The are 16 DNA matches who are in this class, highest is a mere 21cM with Zero shared match’s.
My Paternal and Maternal lines are from very different parts of England so the probability of this happening in my case in recent times is in any case very, very low.
Using the Gedmatch “Are My Parents Related” tool produces a NO.
As I said a few replies ago, your parents don't have to be related for you to have a "both sides" match. You just need to have a cousin on your father's side to marry a cousin from your mother's side and produce a child. That child would be a both sides match to you.
You say the probability in recent times is low. But people were much more mobile in the 20th / 21st century. Servicemen and women moving around in wartime and more recently people moving long distances from their birthplace for jobs.
My mother was in WRAF in WW2. came from very south of Hampshire. Was posted all over UK even as far north as the Orkney Isles. She could easily have met a man and married him in any one of those places. Neither she nor her two sisters married a man from their home town. On the other side of the family, two aunts in the ATS from Berkshire married husbands from Liverpool and Devon, also in the services.
Going further back to 1920 (and this took a bit of finding because the Smith surname was involved) my 1C1R from Portsmouth married a man from Durham. We discovered he had briefly (between censuses) been a sailor (merchant navy) and met her in Portsmouth.
-
I would agree that the majority of the both sides that Ancestry has assigned are spurious. My genuine both sides that I found, and Ancestry didn't, is only 16cm/ 1 segment, unweighted 25, longest 25cM
-
Thanks again all for your info. I am now convinced this update is spurious. I will report as such to Ancestry support. Let you know if I get a response.
-
I too have some both side.
The are 16 DNA matches who are in this class, highest is a mere 21cM with Zero shared match’s.
My Paternal and Maternal lines are from very different parts of England so the probability of this happening in my case in recent times is in any case very, very low.
Using the Gedmatch “Are My Parents Related” tool produces a NO.
As I said a few replies ago, your parents don't have to be related for you to have a "both sides" match. You just need to have a cousin on your father's side to marry a cousin from your mother's side and produce a child. That child would be a both sides match to you.
You say the probability in recent times is low. But people were much more mobile in the 20th / 21st century. Servicemen and women moving around in wartime and more recently people moving long distances from their birthplace for jobs.
My mother was in WRAF in WW2. came from very south of Hampshire. Was posted all over UK even as far north as the Orkney Isles. She could easily have met a man and married him in any one of those places. Neither she nor her two sisters married a man from their home town. On the other side of the family, two aunts in the ATS from Berkshire married husbands from Liverpool and Devon, also in the services.
Going further back to 1920 (and this took a bit of finding because the Smith surname was involved) my 1C1R from Portsmouth married a man from Durham. We discovered he had briefly (between censuses) been a sailor (merchant navy) and met her in Portsmouth.
Yes, but I am not talking about the 20th century, especially with a match of only 21cM as that is probably the result of a joining in the early to mid 19th century.
At that level the highest probability is in the 4C range or at my 3xGGP.
All my 3xGGP except I were fairly static and the one that was not has a lot of descendants but the 21cM has no shared match’s with this 3xGGF line so that skews the though process. If this changes then I shall revisit my views.
I have compared my Wife’s DNA to my own with no shared segments the results and given that within the last four generations there has been no interaction between our families which is not surprising since we are from Cities 70 miles apart.
On Ancestry we do have shared match’s with each other so as you say Cousin marrying Cousin but in our cases that is likely to be 200+ years ago.
-
Why would the marriage event need to be so long ago?
My genuine both sides match is 16 cm, the marriage which joined the two lines took place in 1901. But it could easily have been more recent. The match is my 3rd cousin on my father's side and my 5C1R on my mother's side. I am surprised it is so low a match. The 3rd cousin relationship is by being great grandson of my great grandmother's sister. I have a 34 cM match with a 3C1R, the 2 x great grandson of the same sister, and a 62 cM match with another 3C1R, the 2 x great grandson of another sister of the same great grandmother.
Painter predicts a much more distant relationship for my 16 cm match, more in line with the 5C1R alone. But if it wasn't for a good paper trail and 9 shared matches to descendants of that great grandmother's parents I might doubt the 3C relationship
Isn't DNA wierd ;D
-
I now have 4 Both Sides matches where previously I had none. Two of them share matches exclusively with my paternal side, one exclusively with my maternal side and one with both sides. My mother comes from a well-researched English line that lived and bred in the English Midlands going back to the mid 1700s. My genetic father was an Irish American who went to the USA from Co. Mayo and didn't go to England until the war.
It is very very unlikely that I have both sides ancestors based on the timescales of autosomal DNA accuracy.
-
I've just seen these on my wife's profile. One of the 5 such matches shares a common ancestor on her paternal line, albeit at the 5C level
-
My sister is showing as 50 Both Sides, I’m showing as having 37
I searched my sisters 50 on my list and apart from each other, her daughter and another niece I found 5 we both have, a 20cM a 15, 13,11, and 9. No idea how any of the 5 match on either side.
Our 1st Cousin has 26 both sides I found 1 who matches with me, none with my sister.
-
Descent on both sides is certainly something that occurs but many such matches would probably be too long ago for Ancestry to recognise them. My wife and I are descended from the same couple born 1751 and 1760, my 4 x GGPs her 6 x GGPs, but we don't show as a match to each other (ca 4 cM using GEDMatch).
Perhaps more such cases are when the descent is from a small population (village) where there were a number of events over the years
-
interesting thread. My both sides matches have increased from 4 to 41 in this latest update. Good to see others are bemused by this as well!
-
I was about to ask where the “both sides” matches are to be found, and I have just noticed that a “both sides” has suddenly appeared where there wasn’t one before.
I’ve only ever had Parent 1, Parent 2 and Unassigned. I presume the “both sides” wasn’t showing because I didn’t have any matches in common with both sides, which is as I would expect.
Now two have popped up - one 12cM and one is 9cM.
Ancestors of my Parent 1 and Parent 2 were between 2.000 to 8,000 km apart so any kind of cM match is fairly difficult to believe. :)
I have no common matches with either of them.
-
My "Both Sides" matches have jumped from 2 to 74 in latest update, ranging from 45cM to 8cM.
Not analysed them all yet but seems odd.
-
Ancestors of my Parent 1 and Parent 2 were between 2.000 to 8,000 km apart so any kind of cM match is fairly difficult to believe. :)
It doesn't matter if your two sets of ancestors were widely separated, it just needs a descendant of each set which may be your generation or your parents generation to meet, marry and produce a child who takes a DNA test. My two lines from a couple of hundred miles apart came together in the 20th century to produce ME and to produce my only genuine both sides match. Descendants of your widely separated ancestral lines obviously came together to produce you, so why not another descendant from each branch getting together?.
Ancestry failed to pick up my real both sides match, just makes it paternal because that is the closest relationship - paternal 3C and maternal 5C1R
-
Ummm, still not convinced of accuracy, but understand what you are saying Lizzie. ;D
If Ancestry didn’t pick up your real both sides match, there’s a possibility some of the both sides matches are misassigned too.
I await any new updates to see what they bring, and hope for some matches in common to appear (which might help clarify).
-
If Ancestry didn’t pick up your real both sides match, there’s a possibility some of the both sides matches are misassigned too.
I'm pretty sure most are spurious, but haven't checked thoroughly
-
Between the two of us, my brother and I have 29 new both sides matches. All but one are at match lengths of 10 cM or less. None are shared between us, they are all apparently unique to either him or me. For three of the matches there are shared matches with other matches on our maternal side (2) and paternal side (1). Several of those shared matches are already confirmed.
For one of the matches with shared maternal matches, there is a surname in her tree that matches a surname on our paternal side. I spent a couple of hours researching her line with that name back to c. 1780, but with no indication that her line is connected to ours in any way.
As none of the matches appear to be necessary to corroborate my own tree, and most of the match lengths are in the region where IBC is a distinct possibility, particularly with no other shared matches and none at all being attributed to both of us, I'm not going to spend any more time on them.
-
One thing I have noticed is that all but one of my new "both sides" matches have 2 segments. Given that the matches range from 8 to low teens cMs this is rather unusual, normally at these low values it's a single segment.
-
One thing I have noticed is that all but one of my new "both sides" matches have 2 segments. Given that the matches range from 8 to low teens cMs this is rather unusual, normally at these low values it's a single segment.
It is the same with my two “both sides” matches - 2 segments.
-
Mine has also gone up - from 1 to 12. One I had already identified as having shared matches on both sides of my tree and is a 32cm match but with no tree so I have been unable to get any further.
The rest have no shared matches and range from 8 - 15 cm but they are all on two segments. Intriguing. Six of them have trees of some sort so I may try to investigate further, none of them have shared surnames and the only shared location is London, UK - so it may be tricky!
-
Just received an email from Ancestry informing me that yet more of my matches have been sorted by parent. Having checked, there are no more both parent additions, and I don't keep track of how many matches I have to each parent, but what is noticeable is that I'm being told on the main DNA matches page that my top four maternal common communities are all in the USA, and three out of the top four paternal communities are also either in the USA or Australia.
If I look at the communities assigned on the ethnicity page, they are the same as they have been for some time, all in the UK or NW Europe apart from two minor American settler communities...
-
As mentioned earlier, I have 74 "both sides" matches (previously this was 2).
From checking number of segments for each match:-
- only 1 match has 1 segment (a 15cM match)
- 2 have 4 segments
- 5 have 3 segments
- 66 have 2 segments
So far, I haven't been able to make much sense of these matches.
-
Just received an email from Ancestry informing me that yet more of my matches have been sorted by parent. Having checked, there are no more both parent additions, and I don't keep track of how many matches I have to each parent, but what is noticeable is that I'm being told on the main DNA matches page that my top four maternal common communities are all in the USA, and three out of the top four paternal communities are also either in the USA or Australia.
If I look at the communities assigned on the ethnicity page, they are the same as they have been for some time, all in the UK or NW Europe apart from two minor American settler communities...
Mine look the same for both Parents and still about the same number of ‘unassigned’. No obvious changes with their update.
I noticed yesterday that my “common communities” had changed, though I generally ignore that. I hadn’t noticed Puerto Rico on the list before - that’s an interesting one. :)
Are the “common communities” based on locations in people’s trees? Not many of my DNA matches have trees, which would reflect in the common communities not being as expected? I wonder if more Americans having trees would explain more US locations in common communities.
I’d find it more useful to show matches’ locations (as My Heritage does) than “common communities”.
-
Are the “common communities” based on locations in people’s trees? Not many of my DNA matches have trees, which would reflect in the common communities not being as expected? I wonder if more Americans having trees would explains more US locations in common communities.
Yes, there's an element of Ancestry family tree scouring involved, in that their algorithm looks at clusters of DNA matches and also the locations contained in the matching members' family trees.
I do have a fairly large number of half cousins in the US, all from three sons of my GGF fathered by a different wife - not my GGM. They are by far my largest group of matches derived from emigres, with many descendants having taken Ancestry DNA tests; although I also have some direct cousins in Australia and Canada. Canada isn't mentioned in these new communities at all.
But the main point is that these eight foreign communities are the communities I am most attached to according to the DNA matches by parent page, yet when I drill down to the communities page itself, I'm told that I am connected to The Midlands, Central Southern England, the East of England, Wales and "Ohio, Indiana and East Kentucky Settlers". Irrespective of their accuracy or otherwise (some are broadly correct, others I have an issue with) they completely contradict my "top communities" now shown on the DNA matches by parent page, and which as far as I'm concerned are complete bunkum.
Something definitely seems rotten in the state of Denmark, not that I have a community there either, but if I did I would regard it as far more likely than the ones that I seem to have just been allocated ;D
-
Interesting, I used to describe myself as an Anglo, Welsh, Canadian.
DNA came on the scene and I became:-
Anglo, Welsh, Canadian, Indian, Italian, Scottish, French, German, Spanish, Belgian, Norwegian, Icelandic, Greek, Syrian, Albanian, Hungarian, North Macedonian, Bulgarian, Turkish, Austrian, Lebanese, Palestinian, Israeli & Jordanian.
Then things expanded and added to the above was:-
Luxembourg, Switzerland, a smidgen of the Netherlands, not no mention, Jersey, Guernsey, Sark, Alderney and a new lifeform under a rock on the beach of Brittany.
Ahh such is life
-
The main issue that I have with these North American and Australian communities though is that I am not directly descended from anyone who lived in those places. Rather, some of the brothers, sisters, aunts and uncles (half and full) of my direct ancestors settled in those places for various reasons within the last 200 years.
If I accept Ancestry's description of communities, that they are formed from Ancestry DNA members whose ancestors probably came from the same place of cultural group within the last 50-300 years, it would make sense for my cousins in other continents to be attached to communities in the British Isles. It makes far less sense for me to be attached to communities in other continents where neither I or my direct ancestors have ever resided.
And even less sense for Ancestry to state that those communities in N. America and Australia are the communities to which I have the highest attachment on the DNA matches by parent page, when none of the communities listed there appear in the list on my Communities page itself ???
It has rather reinforced my opinion of the whole concept, at least as far as Ancestry are concerned. This is one aspect for which Living DNA have a far more precise UK biased implementation. They have no access to a tree relating to my DNA sample, yet their recent ancestry results are pretty accurate in comparison to my research, in some places down to county level. Even within the UK, Ancestry seem far more hit and miss.
-
Conversely for me Living DNA is way off the mark.
Summation to 30% for East Anglia, Midlands, Cornwall and Scotland.
I have Zero ancestors from these areas in the last 250 years.
-
What you say makes sense Phil. I think “common communities” is a fairly meaningless and unhelpful feature.
It’s based on where some people who match some of your DNA live now, or can be pigeonholed into particular groups.
(“Last names in trees” has no value either I don’t think.) :-\
With Parent 1’s side I have two lots of Ashkenazi Jewish communities - one in Central and Eastern Europe and another in NE Europe.
Although I only have an estimated “13% Jewish” on Ancestry I get many matches who have very high percentages of Jewish “ethnicity” - some into the 90%s.
I think that they are keen on DNA tests, so are overrepresented in the number of matches I have, and this must be the reason for my two Jewish common communities.
Unfortunately my Puerto Rican community will probably remain a mystery. ;D
Parent 2’s side is questionable too, but I won’t elaborate.
-
Conversely for me Living DNA is way off the mark.
Summation to 30% for East Anglia, Midlands, Cornwall and Scotland.
I have Zero ancestors from these areas in the last 250 years.
Is that on a transferred in kit Biggles? The results for my Ancestry DNA file transferred to LDNA are quite vague. The results on the test I took with LDNA are much more specific.
Of course, the other possibility is that I am a victim of the expectation bias that I always warn about, and your results confirm that the "science" is as imprecise as I generally believe!
-
(“Last names in trees” has no value either I don’t think.) :-\
I use that feature from time to time, and quite like it. Yes, it can turn up a lot of useless results, but especially with less common names, I have had some success in identifying connections to my tree where I had otherwise drawn a blank, including a tree that I almost discounted because all the information led to the completely opposite direction of England. But when I went through it more closely, I realised that the owners Gx3GF had been baptised in a location of interest to me. It ended up unlocking a further two generations for me after a lot more work.
-
(“Last names in trees” has no value either I don’t think.) :-\
I use that feature from time to time, and quite like it. Yes, it can turn up a lot of useless results, but especially with less common names, I have had some success in identifying connections to my tree where I had otherwise drawn a blank, including a tree that I almost discounted because all the information led to the completely opposite direction of England. But when I went through it more closely, I realised that the owners Gx3GF had been baptised in a location of interest to me. It ended up unlocking a further two generations for me after a lot more work.
Parent 2 has three familiar surnames (all related to my Irish) and others which are a mystery. Parent 1 = no familar surnames.
I think I’ve been unlucky with matches lack of trees and reluctance to reply to messages. :(
-
Conversely for me Living DNA is way off the mark.
Summation to 30% for East Anglia, Midlands, Cornwall and Scotland.
I have Zero ancestors from these areas in the last 250 years.
Is that on a transferred in kit Biggles? The results for my Ancestry DNA file transferred to LDNA are quite vague. The results on the test I took with LDNA are much more specific.
Of course, the other possibility is that I am a victim of the expectation bias that I always warn about, and your results confirm that the "science" is as imprecise as I generally believe!
Yes it is Phil.
I do have a Living DNA test to undertake so it will be a few weeks before they give me the results.
The My Heritage site has both their own and Ancestry’s DNA data, and that shows 36% Scandinavian which is way off the mark but apparently typical of their data
-
I have both sides matches in ancestry as well - around 400. In my case I know that there are quite a few about, including several in the present day which I had always known about.
I had two cousins from each side test their DNA some years ago and I found that they were all related - that is despite coming form either my mother's or father's side as a very close match to me they also matched to the other parent's side at around 5 - 8 cousins. I uploaded to gedmatch and used the one that shows generations and I have close generation match with them as it should be but the other's are related to each other at a distant generation.
Living in our area I have found that 99% of our ancestors come from the same isolated area mostly from around a 50 mile diameter circle and if that's the case I'm quite sure that there was a lot on inter marriage, even today you hear about 3rd or 4th cousins marrying and believe me they know who's who because we all know each other's family having grown up with our parents saying that so and so is related, but with no direct line to see.
Thank you - sorry I waffled
Nesta
-
Yes it is Phil.
I do have a Living DNA test to undertake so it will be a few weeks before they give me the results.
It will be interesting to see if you find any difference, good or bad
-
Parental assignment employs a computer algorithm that utilizes known DNA matches and scans Ancestry family trees to identify lines of descent and correlate them across trees, thereby linking to families. Similar to Thrulines, it may be prone to errors if user trees contain inaccuracies, and it is currently labeled as being in beta phase. Notably, it does not solely rely on an analysis of your DNA. As autosomal tests are not phased, the phasing process is estimated.
Ancestry acknowledges that the accuracy of parental assignment will be at least 90% for approximately 9 out of 10 users, while being less than 90% accurate for the remaining 1 in 10.
-
I have a match with 42cM across 3 segments. She has the same uncommon surname as my mother. She has been correctly identified by Ancestry as having common ancestors with me (our 2 x great grandparents). She is definitely my 3rd cousin on my mother's side. Our trees on Ancestry overlap. Our four shared matches have all been identified correctly by Ancestry as being maternal. But Ancestry designates her as "unassigned". How does the algorithm work that one out?
-
I have a match with 42cM across 3 segments. She has the same uncommon surname as my mother. She has been correctly identified by Ancestry as having common ancestors with me (our 2 x great grandparents). She is definitely my 3rd cousin on my mother's side. Our trees on Ancestry overlap. Our four shared matches have all been identified correctly by Ancestry as being maternal. But Ancestry designates her as "unassigned". How does the algorithm work that one out?
I have a similar match (DN) where our common ancestor is correctly identified by Anc but sideview technology has not assigned the match to either parent. Interestingly another match (CS) who has the same common ancestor has been correctly assigned. All three of us are on MH and the chromosome browser shows my match with CS as triangulated over three segments but only one with DN. Think that suggests the lack of assignment for DN is because the sideview technology does not consider the match sufficient to assign to a parent with sufficient confidence. It does highlight that sideview technology does not use trees in making the assignment - only DNA is used. I am guessing your situation is similar?
-
Yes, as I understand it, the situation is as ikas says.
The segments that your match has in common with you don't contain enough information for Sideview to make a determination based on what it currently knows about the segments of your DNA already assigned to one side or the other, as the information available from that match is not sufficiently in common with your existing assigned matches. If you are certain about the parental side of the match, you can manually assign it and that should help improve the Sideview algorithm for you going forward.
-
Interesting discussion, I'd like to add my thoughts.
Like many on here my both sides matches jumped from just my sister to about 40, all of which significantly have 2 segments that match.
It seems to me that the phasing falls into two stages, first separate out the base pairs into phased segments, then allocate those segments to a parent. The first part should be relatively easy given sufficient matches, the second part much harder as it would depend on matches having two or more segments in common. If this is correct then the errors are likely to come in at the parental allocation stage.
So then what I see as the underlying data that Ancestry have calculated would be a bunch of phased segments allocated to one parent or other, say for arguments sake about two hundred. Of these two hundred segments 90% are correctly assigned to a parent and 10% (about 20 segments) wrongly (based on Ancestry's numbers).
So then to get a spurious both sides allocation it is only neccesary to match within one of the segments in the correct group and one of the segments in the wrong group. Hence my both sides matches all have 2 segments in common.
Quite why the number of both sides matches has recently jumped I have no idea, presumably changes in the algorithm.
An interesting side effect of this process would be the ability of Ancestry to eliminate a large amount of IBC matches since they would match neither set of parental phased data.
-
If only Ancestry bought out Gedmatch and integrated their tools into Ancestry!
With my Both Sides being below 25cM they are not really worth researching.
Same with my Wife’s matches, all are lowish cM.