Page 1 of 3 123 LastLast
Results 1 to 10 of 24

Thread: DNA matches - another format change

  1. #1
    Registered Users
    Posts
    1,349
    Sex
    Location
    Brisbane
    Nationality
    Australian
    Y-DNA (P)
    T-P322 (T1a2b1)
    mtDNA (M)
    H6a1

    Australia Cornwall England Scotland Germany Poland

    DNA matches - another format change

    This morning after clicking on DNA matches up came a selection of tabs:
    All matches By Parent BETA By Ancestor By location

    "By ancestor" is simply Thru Lines (TM) re-badged for newbies (and if it helps them help me, why not?)

  2. The Following 4 Users Say Thank You to Saetro For This Useful Post:

     Dts (01-13-2023),  JMcB (12-08-2022),  Mr.G (12-07-2022),  msmarjoribanks (12-19-2022)

  3. #2
    Registered Users
    Posts
    162
    Sex
    Location
    TX
    Ethnicity
    English, German, Scottish
    Nationality
    American
    Y-DNA (P)
    R1b-U106-Z8-S1774
    mtDNA (M)
    T1a1p
    Y-DNA (M)
    R-U152

    England Scotland Germany Netherlands
    I have noticed another little tweak. At least I think its new.

    When one looks at a set of Shared Matches and then selects the Groups tab, those Custom Groups (dots) that do not point to any further matches are grayed out.
    Last edited by MAbrams; 12-08-2022 at 01:53 PM.

  4. The Following 2 Users Say Thank You to MAbrams For This Useful Post:

     JMcB (12-08-2022),  Saetro (12-08-2022)

  5. #3
    Global Moderator
    Posts
    1,627
    Sex
    Location
    Cambridge MA / Rome, Italy / San Diego, CA (currently)
    Ethnicity
    Polish/British Isles
    Nationality
    U.S.
    Y-DNA (P)
    R-Y154732
    mtDNA (M)
    H1
    mtDNA (P)
    J1c2

    Poland England Ireland Munster
    Quote Originally Posted by Saetro View Post
    This morning after clicking on DNA matches up came a selection of tabs:
    All matches By Parent BETA By Ancestor By location

    "By ancestor" is simply Thru Lines (TM) re-badged for newbies (and if it helps them help me, why not?)
    When I select "by Parent" I get matches with no Common Ancestor. Separatly "Common Ancestor" matches and the matches appearing on Thru Lines have always appeared to be the same. So "By Parent" and Thru Lines do not appear to be the same to me.
    Last edited by pmokeefe; 12-08-2022 at 04:43 PM.
    YFull: YF14620 (Dante Labs 2018)

  6. The Following 2 Users Say Thank You to pmokeefe For This Useful Post:

     JMcB (12-08-2022),  Saetro (12-08-2022)

  7. #4
    Registered Users
    Posts
    1,806
    Sex
    Location
    Pittsburgh, PA, USA
    Ethnicity
    Ger.-Brit.-Catalan-more
    Nationality
    (U.S.) American
    Y-DNA (P)
    R-YP619*
    mtDNA (M)
    H1bg

    United Kingdom Germany Bayern Catalonia France Ireland Switzerland
    Quote Originally Posted by pmokeefe View Post
    When I select "by Parent" I get matches with no Common Ancestor. The "Common Ancestor" matches and the matches appearing on Thru Lines appear to be the same. So "By Parent" and Thru Lines do not appear to be the same to me.
    What I was expecting when I saw "By Ancestor" among the ways of selecting matches -- "All Matches"; "By Parent"; "By Ancestor"; "By Location" -- was that Ancestry was attempting to do the same sort of thing I've already been doing. I created an Excel file in which I've been assigning my matches by grandparent or even earlier ancestor when possible. Only it really is just ThruLinesR,

    This is pretty disappointing, because ThruLinesR relies on matches having trees and doesn't really take advantage of Ancestry's SideViewTM. What I'd hoped for was that Ancestry would begin grouping matches the way I do, which is by grandparent and even great grandparent or earlier ancestor when possible.

    For example, if I click on my paternal grandfather's name in ThruLinesR, only seven names show up as clickable links to matches. I'll ignore the first two, since they're my daughter and the daughter of one of my full sisters. The other five are a 1st cousin (VE), three 1st cousins once removed (JS, SB, and ZH), and a 1st cousin twice removed (BH).

    However, I actually have a lot more paternal 1st cousins and 1st cousins once removed than this. I have three more 1st cousins and six more 1st cousins once removed on this side. I'll exclude one of the 1st cousins once removed since she isn't descended from my grandparents, but my great grandparents. She's my grandfather's niece.

    It seems to me that Ancestry ought to be easily able to assign these matches to my grandfather even in instances in which they can't determine the exact relationships, although in some cases they should be able to do just that. For example, Ancestry shows my 1st cousin JS as my grandfather's grandson through one of his sons. This is the same son who is the father of my 1st cousin VE, so VE and JS are aunt and nephew. But in this case my 1st cousin DS is also one of my matches at Ancestry, and happens to be a sister to VM and mother to JS. Ancestry absolutely should be capable of identifying parent-child and sibling relationships easily.

    For one thing, parents will always show half matching with their children across the entire length of all autosomes. (They may have noticeably more matching if the the parents are closely related to each other; for example, if they're 1st cousins.) Full siblings will also be obvious, although half siblings will be less so. In any case, it should be possible for Ancestry to at least determine that DS is another descendant of my paternal grandfather -- and of my paternal grandmother, for that matter.

    Then there's my 1st cousin once removed, CH. CH is a brother to ZH, who is shown as one of my ThruLinesR matches under my grandfather. And there's RD, who is the daughter of VM above.

    I believe Ancestry could assign many more relatives to ThruLinesR ancestors -- not necessarily in their exact relationships, but at least as likely descendants. I do the same thing already, but based on shared matches. Obviously I can't compare DNA directly since Ancestry continues to refuse to provide any sort of chromosome browser, but Ancestry itself does not have this limitation since they have to be able to determine matching in the first place.
    Besides British-German-Catalan, ancestry includes smaller amounts of French, Irish, Swiss, Choctaw & another NA tribe, possibly Catawba. Avatar picture is: my father, his father, & his father's father; baby is my eldest brother.

    GB

  8. The Following 2 Users Say Thank You to geebee For This Useful Post:

     msmarjoribanks (12-19-2022),  Saetro (12-08-2022)

  9. #5
    Registered Users
    Posts
    1,806
    Sex
    Location
    Pittsburgh, PA, USA
    Ethnicity
    Ger.-Brit.-Catalan-more
    Nationality
    (U.S.) American
    Y-DNA (P)
    R-YP619*
    mtDNA (M)
    H1bg

    United Kingdom Germany Bayern Catalonia France Ireland Switzerland
    There are other ways in which it doesn't make sense for Ancestry to simply offer ThruLines as a way of showing matches "By Ancestor" here. For example, I've noticed that my 1st cousin once removed, "SB", is shown as one my matches under my paternal grandmother in ThruLinesR, but not as one of my matches under my paternal grandmother's father.

    This is because SB made a mistake in her tree. I've sent her a message about it but she hasn't corrected it yet. The error wasn't entirely hers, since she followed a the presumption made by whosever wrote the summary of the 1920 census that included my grandmother.

    The summary gives a "father's name" for my grandmother, but it's an illustration of why one shouldn't "assume". Obviously the person who summarized the census assumed that if my grandmother -- who has the same surname as the head of house -- is the grandmother of the head of house, then she must be the daughter of one of that woman's sons.

    Of course, how does he know that she's the daughter of any of the folks named here? He -- or she -- doesn't know this. My grandmother could have been living with her grandmother because she was an orphan in the care of her grandmother Or she could have been the daughter of the other son listed.

    In fact, my grandmother was neither of these. She was born out-of-wedlock and not given her father's surname. But I've actually seen my grandmother's birth certificate, which names both of her parents even though she wasn't given her father's last name.

    As it turns out, the census simply names, first, the head of house -- my grandmother's grandmother -- then it names her daughter (my great grandmother), two of her sons; and lastly, my grandmother. They're shown in age order -- and possibly also generation order -- with my grandmother of course being last. The man listed immediately before her isn't in that spot because he's her father, as the summarizer obviously assumed; but simply because he was the youngest offspring of the head of house, as well as older than the granddaughter.

    In addition to having seen my grandmother's birth certificate -- and the fact that it was never a secret who my grandmother's father was -- I have several half 2nd cousins who are descended from my grandmother's father. For example, there's AW, who shares 179 cM with me. There's also JB, who shares 160 cM; KH, who shares 136 cM; and FZ, who shares 128 cM.

    The most significant thing, though, is that several of these half 2nd cousins once removed are also in the shared match list between SB and me. So they're not only related to me but to her, and although I can't tell how much she shares with them, I think it's likely to be in an amount consistent with their being half 2nd cousins once removed to SB.

    SB simply made a mistake in her tree, but Ancestry has to know that both trees can't be right here. My grandmother could not have had two different biological fathers. (Of the nonbiological sort she had none.)

    The DNA -- and presumably SideViewTM -- almost certainly would support SB's inclusion as a descendant of my grandmother's father, and the SideviewTM also supports that this man was not the one mistakenly claimed by SB, but the one from both SB and I share descent with many other matches.
    Last edited by geebee; 12-08-2022 at 06:20 PM.
    Besides British-German-Catalan, ancestry includes smaller amounts of French, Irish, Swiss, Choctaw & another NA tribe, possibly Catawba. Avatar picture is: my father, his father, & his father's father; baby is my eldest brother.

    GB

  10. The Following User Says Thank You to geebee For This Useful Post:

     Saetro (12-08-2022)

  11. #6
    Global Moderator
    Posts
    1,627
    Sex
    Location
    Cambridge MA / Rome, Italy / San Diego, CA (currently)
    Ethnicity
    Polish/British Isles
    Nationality
    U.S.
    Y-DNA (P)
    R-Y154732
    mtDNA (M)
    H1
    mtDNA (P)
    J1c2

    Poland England Ireland Munster
    When I select "Parent Matches" I get some with No Trees (42 out of my top 200 parental matches, for examples).
    YFull: YF14620 (Dante Labs 2018)

  12. #7
    Registered Users
    Posts
    1,806
    Sex
    Location
    Pittsburgh, PA, USA
    Ethnicity
    Ger.-Brit.-Catalan-more
    Nationality
    (U.S.) American
    Y-DNA (P)
    R-YP619*
    mtDNA (M)
    H1bg

    United Kingdom Germany Bayern Catalonia France Ireland Switzerland
    Quote Originally Posted by pmokeefe View Post
    When I select "Parent Matches" I get some with No Trees (42 out of my top 200 parental matches, for examples).
    So do I, when selecting "By Parent". Not when selecting "By Ancestor". But parents are ancestors, too, and why should the behavior be any different whether it's "By Parent" or "By Ancestor"? It's just a matter of narrowing the "filter". "By Parent" encompassing all of the ancestors of my father -- my paternal side -- and all of the ancestors of my mother -- my maternal side.

    Truly being able to select by, say, a certain grandparent or great grandparent should give you matches limited to just that person's side. Only right now selecting "By Ancestor" does nothing more than provide a short cut to ThruLines -- even if the ancestor in question is one of the parents!

    I'm hoping that eventually selecting "By Ancestor" will mean just that -- that it won't simply be ThruLines but will actually attempt to cluster matches who share that ancestor, including some matches without trees or whose trees are contradictory at that point.

    That's why I mentioned my 1st cousin once removed, SB, in my earlier post. SB does show up as one of my matches when I click on the name of either of my paternal grandparents, because our trees are in agreement up to that point. But if I click on the name of my grandmother's mother, SB no longer shows up as a match. All other matches still show up, but since SB has the wrong person listed as my grandmother's father, Ancestry excludes her.

    But as I stated earlier, my grandmother can't have more than one biological father! And in this case, the question for Ancestry is, which one has not only the support of a tree but also the support of DNA? It isn't that descendants of my great grandfather don't show up as shared matches with SB; at least some of them do.

    It's also abundantly obvious that full siblings have identical trees -- at least as far as ancestors are concerned. So if you have two people with DNA indicative of their being full siblings and only one has a linked tree, then the linked tree is applicable to both of them as long as it's correct.

    So, for example, if one of my matches is a full paternal 1st cousin to me -- meaning we share both paternal grandparents -- then a sibling of that match is also a full paternal 1st cousin to me. So if one shows as a match under either of these grandparents, so should the other.

    Unfortunately, this is not what "By Ancestor" means at present. As I've said, it's just a short cut to ThruLines. But it could be more, and I'm hoping eventually that it will be more.
    Last edited by geebee; 12-09-2022 at 12:51 AM.
    Besides British-German-Catalan, ancestry includes smaller amounts of French, Irish, Swiss, Choctaw & another NA tribe, possibly Catawba. Avatar picture is: my father, his father, & his father's father; baby is my eldest brother.

    GB

  13. The Following User Says Thank You to geebee For This Useful Post:

     pmokeefe (12-09-2022)

  14. #8
    Registered Users
    Posts
    1,806
    Sex
    Location
    Pittsburgh, PA, USA
    Ethnicity
    Ger.-Brit.-Catalan-more
    Nationality
    (U.S.) American
    Y-DNA (P)
    R-YP619*
    mtDNA (M)
    H1bg

    United Kingdom Germany Bayern Catalonia France Ireland Switzerland
    Quote Originally Posted by MAbrams View Post
    I have noticed another little tweak. At least I think its new.

    When one looks at a set of Shared Matches and then selects the Groups tab, those Custom Groups (dots) that do not point to any further matches are grayed out.
    I'm not entirely sure what you mean, but to my annoyance matches that I have indicated a side for are not shown even when I select that side. They're only shown when I select "custom labeled" for that side.

    This includes a paternal 1st cousin once removed, even though her mother is shown as one of my paternal 1st cousins. If the mother is paternally related to me, so is the daughter.

    Of course, it's possible that her DNA test results are too recent, although I'm remembering them as having been available some time ago. I checked my cousin's "join" date, and it's 2017. Unfortunately, that isn't really very useful because it doesn't mean she tested then.

    Another possibility is that the daughter in this instance is also related to me (through her father) on my mother's side. I really doubt it, but it's possible. However, even if that were true, she would still clearly be a paternal match. She would simply also be a maternal one.

    The problem is, it's obvious that Ancestry doesn't understand the meaning of "side". If they did, they would not have told me this of a match I am convinced is strictly maternal, but which they believe is "both":

    We think this match is on your Both side, but you labeled them Maternal. Do you want to keep your label or update it?

    First, we only have two sides -- because we only have two biological parents. We may have people that are related on one side and the other, but it doesn't put such people on a "third" side. It certainly doesn't mean "Both" instead of either, it just means you're be correct no matter which side you say it's own.

    With my daughter, for example, she is indeed on "Both" sides. But she isn't on both sides instead of being on my father's side, or instead of being on my mother's side. She's on both sides because she is on my father's side and she is on my mother's side.

    Despite this fact, what Ancestry does if you select "Maternal" is to only include those who are purely maternal. Likewise for the selection of "Paternal". That means they omit some matches which do belong to the side you're looking at, simply because they either do -- or Ancestry believes they do -- happen to belong to the other side as well.

    I consider this to be similar to a situation in which someone surveys dog owners and cat owners. Perhaps because you're allergic, you want to know who owns cats. But on the survey, not all cat owners are reported. Why? Because they're also dog owners. But you allergy doesn't care!

    Likewise, maybe you're interested in seeing everyone who is maternally related to you. Anyone who is also paternally related to you -- no matter how distantly, as long as there's the smallest detectable sharing on that side -- will not be reported as a maternal match, even though that's what you wanted to know about.

    To return to my maternal match that Ancestry insists is a "Both" match, their statement to me gives the impression that I'm wrong. In actuality, they're saying I'm right. It is a maternal match. Their only quibble with me is that they think it isn't only a maternal match, even though they provide absolutely no evidence to support that.

    Out of nearly two hundred shared matches, every last one of them is on my mother's side. Of course, this doesn't mean there might be someone who shares less than 20 cM with this match, since that's the threshold for inclusion on the shared match list. It does mean, however, that they're much more likely to be a maternal match than a paternal match.

    If Ancestry wants me to believe she's both, I want more evidence. Why? Because I know that while SideView

    Could there be a shared match that is paternal but simply too small to report? Yes, since there's a reporting threshold of 20 cM in order to report shared matches as such. But since Ancestry actually agrees that this is a maternal match -- even if they think it's also a paternal match -- they shouldn't hassle me without offering any sort of evidence except their say-so.

    The reality is this. While SideViewTM may do a pretty good job at phasing -- even in the absence of having tested parents -- it isn't perfect. It is very likely that some bits of DNA end up being assigned to the wrong side, which might affect the determination of side in some cases.

    This "misattribution" of DNA happens even when people have two tested parents, so it will certainly sometimes happen with SideViewTM, which is why I don't want to just take Ancestry's "say-so". A little more evidence would be nice.
    Last edited by geebee; 12-09-2022 at 04:05 AM.
    Besides British-German-Catalan, ancestry includes smaller amounts of French, Irish, Swiss, Choctaw & another NA tribe, possibly Catawba. Avatar picture is: my father, his father, & his father's father; baby is my eldest brother.

    GB

  15. #9
    Registered Users
    Posts
    162
    Sex
    Location
    TX
    Ethnicity
    English, German, Scottish
    Nationality
    American
    Y-DNA (P)
    R1b-U106-Z8-S1774
    mtDNA (M)
    T1a1p
    Y-DNA (M)
    R-U152

    England Scotland Germany Netherlands
    Quote Originally Posted by geebee View Post
    I'm not entirely sure what you mean, but to my annoyance matches that I have indicated a side for are not shown even when I select that side. They're only shown when I select "custom labeled" for that side.
    No Problem...Try #2

    Screenshot 2022-12-09 164614.jpg

    Notice that my Berkey group is very easy to read and has 163 matches
    The other three groups have zero matches and are much fainter or dimmed out.

    ~~~~
    I agree that a Parental filter should show both the Ancestry designated parental side AND any matches you have placed in the corresponding Custom Parental group. Seems odd not to combine them. But this is still a Beta feature so maybe that will be addressed as this product evolves.

    I am not investing any time in assigning to Custom Parental groups as I expect Ancestry will eventually phase these matches.

  16. The Following User Says Thank You to MAbrams For This Useful Post:

     msmarjoribanks (01-04-2023)

  17. #10
    Registered Users
    Posts
    162
    Sex
    Location
    TX
    Ethnicity
    English, German, Scottish
    Nationality
    American
    Y-DNA (P)
    R1b-U106-Z8-S1774
    mtDNA (M)
    T1a1p
    Y-DNA (M)
    R-U152

    England Scotland Germany Netherlands
    Quote Originally Posted by geebee View Post
    I'm not entirely sure what you mean, but to my annoyance matches that I have indicated a side for are not shown even when I select that side. They're only shown when I select "custom labeled" for that side.
    No Problem...Try #2

    Screenshot 2022-12-09 164614.jpg

    Notice that my Berkey group is very easy to read and has 163 matches
    The other three groups have zero matches and are much fainter or dimmed out.

    ~~~~
    I agree that a Parental filter should show both the Ancestry designated parental side AND any matches you have placed in the corresponding Custom Parental group. Seems odd not to combine them. But this is still a Beta feature so maybe that will be addressed as this product evolves.

    I am not investing any time in assigning to Custom Parental groups as I expect Ancestry will eventually phase these matches.


    ~~~~~~~~~~~~~~~


    Quote Originally Posted by geebee View Post
    This includes a paternal 1st cousin once removed, even though her mother is shown as one of my paternal 1st cousins. If the mother is paternally related to me, so is the daughter.
    I have the same situation with a paternal '2nd' cousin once removed, even though 'his' mother is shown as one of my paternal '2nd' cousins. If the mother is paternally related to me, so is the 'son'..

    I am still rather annoyed at this inconsistency. However, when I plot the mother and son (via FTDNA) on DNA Painter, I see that the son, while still matching me for over 100 cM, does tend to place segments in the quieter areas of my chromosome map. He does not line up that well with my known paternal matches. The mother, OTOH, has s few more segments than her son and a couple of them land in my known paternal hotspots. Maybe Ancestry got it right

    To counter against my own argument, you are talking about 1st 1r, with a lot more segments than my 2nd 1r. It does seem hard to fathom that Ancestry could not find suitable segments to phase with other known paternal matches. This assumes of course that the daughter tested before April, 2022.


    ~~~~~~~~~~~~~~~~~~~~



    I find the Ancestry Phasng to be extremely accurate. My high praise does not extend to the Both Sides.

    From the New Phasing Thread. Oct 6. Message #14.
    Quote Originally Posted by MAbrams View Post
    Regarding "Both Sides"

    I have three Purple Bucketed "Both" matches at FTDNA
    All three are at GEDmatch where they show up with "People who match both kits, or 1 of 2 kits", using my two phased kits.

    All three are at Ancestry, and all three are Unassigned. In baseball parlance, a strike out. Or end of inning.
    Heck, I would think there should be at least a maternal or paternal phasing.



    I do have one "Both Sides" at Ancestry. But only at Ancestry.
    Last edited by MAbrams; 12-09-2022 at 11:29 PM.

Page 1 of 3 123 LastLast

Similar Threads

  1. Ancestry's new format
    By Saetro in forum AncestryDNA
    Replies: 13
    Last Post: 01-24-2022, 12:27 AM
  2. Conversion of 23andme V5 to V3 format VCF
    By Ayetooey in forum General
    Replies: 0
    Last Post: 04-05-2021, 07:17 PM
  3. New Raw DNA File Format
    By JimB in forum FTDNA
    Replies: 60
    Last Post: 08-21-2019, 07:55 AM
  4. .txt file format to .fasta
    By [email protected] in forum Inquiries Corner
    Replies: 2
    Last Post: 08-09-2019, 09:25 PM
  5. New format for my aDNA compendium?
    By Jean M in forum General
    Replies: 7
    Last Post: 04-10-2013, 08:30 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •