A Segment-ology TIDBIT
I was adjudicating a ThruLines from a Common Ancestor (CA) down to a Match. The grandchild of the CA didn’t fit. I find about 5% of my ThruLines are wrong so I just dotted the Match yellow (TL Wrong) to add it to that group. But as I was about to close out the Match, I clicked on Shared Matches (which I usually do anyway). The Match was at 13cM so I didn’t expect much. Surprise – over 20 Shared Matches, and almost every one was confirmed or “likely” to be on the line indicated by ThruLines! A clear consensus. I went back to the Match’s line and found another path that worked – back another generation from the ThruLines CA hint!!
The details don’t matter. The moral of this story is that a ThruLines CA AND a consensus of Shared Matches AND the AncestryDNA “side” should all be in agreement. This applies to CAs at other companies, too – the clues should be in agreement.
Takeaways:
1. When you find a CA, be sure to also review the Shared Matches and the side.
2. When you are searching for a CA with a Match, review the Shared Matches first to see if there is a consensus clue.
PS: this assumes you have diligently done your homework and put all known or likely CAs in the appropriate Notes (same for every company).
[22BZ] Segment-ology: Concensus TIDBIT by Jim Bartlett 20231206
segment-ology 
Ciao jim.ho tre persone che ho una triangolazione sul chr 12 che e incorniciata sia dal lato sinistro sia lato destro.anche loro hanno origini dell est.i segmenti sono 8,1cm 7,5cm 8,8cm.ma la triangolazione con tutte e tre e 6,4cm con 5.248 snp.poi se tolgo uno di loro rimangano due il triangolato e di 6,7cm con 5.376snp.che significa?
LikeLike
Kevin – I don’t know. At such low cMs (under 7cM) the chance that they are false segments are very high. For this reason, I do not recommend using using segments under 7cM for Triangulation – too many times it may lead to false results. See the segment diagram in my most recent post – some segments between two crossover points are closely related to an adjacent segment, and sometimes they are not. Jim
LikeLike
Kevin.ciao jim se faccio fare test ha mia madre che ha 74anni posso scoprire il suo lato materno della mia nonna e mia bisnonna?per vedere se qualche corrispondenze che sono sconosiute sono da parte matetna.e per vedere se alcuni match corrispondono ha lei.tu che dici?
LikeLike
Kevin, Testing your mother is a good idea. You should get some clarity. But nothing is guaranteed. Jim
LikeLike
It’s been my experience too.
Most of my incorrect ThruLines have been a generation further on and in an adjacent line, and the Shared Matches have solved the connection.
But my biggest current problem is one where ThruLines has gone right on beyond the generation where the connection probably lies. Mainly because the available trees have some difficult records to deal with. And have made some baffling and illogical decisions.
I am currently waiting for a few more Shared Matches to solve it.
LikeLiked by 1 person
Christopher – thanks for the feedback – I have had some of the same issues. I remain happy to have about 95% of ThruLines correct – and some with records I didn’t have. Putting these lines (down to the Match) in my Tree has resulted in many more ThruLines, I believe. Jim
LikeLiked by 1 person
Hi Jim,
I was wondering if there is elaborate information with regard to size of shared DNA and distance/generations to common ancestors.
For example, would a shared segment of 13 cM between two people be realistically coming from ancestors of 10-13 generations back?
What for segments of 7 cM? What is the chance that common ancestors are 15 generations bacK?
What in case the common ancestor is 12 generations back for 1 person, but 15 generations back for the other person, could a 10 cM fragment then be realistically be shared between these two people?
Thanks in advance for the clarification and sharing your insights
Ben
LikeLike
Ben, see my other reply. Below 20cM almost every mix is possible – type the cM into DNA Painter to see a full range. Jim
LikeLike
Ciao jim sono kevin ho segmento di quel gruppo e poi ho un srgmento che credo sia paterno e quello patetno finisce quando comincia il segmento di quel gruppo sconosciuto che significa che puo essere paterno? Il 7 dic 2023 11:22 AM, “Kevin Capodivento” capodiventokevin6@gmail.com ha scritto:
LikeLike
Keven, Your maternal Chr and your paternal Chr have no knowledge of each other. They were created in two separate human beings. It would be very unusual for a paternal segment to end and a maternal segment to start at the same location. So usually, when you have a Shared Segment (or TG of several segments), it ends at a Recombination Crossover point. That is a point where your father, or some other paternal Ancestor, recombines his or her two Chromosomes (at specific Crossover points) and passed a new Chromosome on to a child. If a segment stops at a Crossover point that was created by your father (from his parents/your grandparents DNA), then the next segment would always be the other grandparent (from your point of view). HOWEVER, if the Crossover was created by your Great grandparent, between his/her parent’s DNA, then the next segment would be from the other Great great grandparent. Long story short, you cannot tell by just looking at the next segment and know which ancestor it was from. It’s from a different Ancestor, but we cannot tell which generation. Sorry, that’s just the way it is. You have to figure out each segment through genealogy. Jim
LikeLike
I have found several instances where Thrulines has connected to the wrong child of a couple, I just keep digging until I know my tree is correct, make a note and go on! i like your idea of using a yellow dot to indicate that it is the wrong line!
LikeLike
The dots have totals, which let me estimate how well ThruLines does. Initially it was over 95% correct. But it’s been slipping the past several months. It’s important for us to try to find the correct links and add them to our Tree. I’ve seen a few incorrect ThruLines removed, a very few corrected (usually when the Match makes the correction, too. A few ThruLines are supported by only one Tree – getting that one fixed – quickly – would help. I’ve actually seen a message before that the path was “under construction” which might change the TL hint (or maybe that was a MyHeritage)
LikeLike
Hi Jim – What do you mean by “the side” in this post? I appreciate everything you have shared and try to understand it, and you have taught me quite a lot and I thank you.
Nancy Leonard
>
LikeLike
Nancy – “side” means parent or paternal vs maternal. I comes from two observations: 1: a vertical Tree has a left side and a right side; and 2. our atDNA has two of each chromosome one on (or from) the maternal side and one from the paternal side. Hope this helps.
LikeLike
Specifically at AncestryDNA, they have designated many of our Matches to be P1 or P2 (with P1 always being on one side or the other – maternal or paternal. If you can figure out which is which, you can have those designations change to Maternal or Paternal (sides) or sometimes “Both” or “None”. These are very beneficial and apply to Private Matches and Matches without Trees – valuable clues…
LikeLike
Ciao sono kevin grazie quel gruppo che ti parlo sempre lato destro 16 che non so da che lato e poi quel gruppo che so che deve essere paterno un segmento del diciamo paterno finisce stessa direzione che comincia il segmento di quel gruppo sconosciuto.ma non triangolano diciamo il srgmento che so che e paterno finisce quando incomincia il segmento del gruppo sconosciuoto ma non triangolano.e possi ile che puo essere paterno e materno?
LikeLike
Funny you should mention this Jim, as only yesterday I had a similar experience. In my case it was a 6C1R. I’m keen to see how long it takes to correct, now I have the correct paths in my tree.
LikeLiked by 1 person
I, too, put the correct line of ancestry from the Match to the different CA. I’m not sure if AncestryDNA will “correct” their info – I have several of these and haven’t seen a change in the ThruLines. It’s us talking to their algorithm – I’m not sure it notes the feedback… Jim
LikeLike
Probably depends on how many people copy me. If the majority follow the incorrect suggested line then it may never update. I live in hope!
LikeLiked by 2 people
This has also been my experience.
LikeLiked by 1 person
Great, I appreciate the feedback.
LikeLike