AncestryDNA’s new BETA

That answers the query I also had, but hadn’t asked. Thank you.

 

In the ThruLines help offered by Ancestry (via the question mark top right of any ThruLines page), one of the questions is 'ThruLines has found potential ancestors that are not related to me. Can I choose not to see these people?' to which the answer is basically to a) 'contact the family tree owner associated with that person to correct any errors or make changes', and b) 'review relevant records and family trees to determine if there actually is a relationship'. How you can do this when you have a potential ancestor where 'evaluate' does not tell you which tree(s) or records the information comes from, I don't know. I have several 'Private' potential ancestors where this is the case.

 

'Private' being the operative word and that along with your (a) answer.."to make contact with that (Private) person..usually with the tag "and ask nicely", is sadly, more often than not, a complete waste of time.

No wonder that I have always been an advocate of Public Trees! At least their owners (given they are at least partly active in the administration of their Trees, and agreed not all of them are) -and again more often than not - do respond, even if we end up having differing opinions when all cards are on the table.

 

Although, there is a point to some of the "potential" matches - as Peter proved when he used an incorrect potential ancestor in ThruLines to break down a brick wall.

 

I agree that the response rate is often disappointing, but my point is how can you even try to contact the private tree owner if ThruLines doesn't tell you what tree the potential ancestor comes from?

I've found that Public tree owners often don't respond either, especially if their tree is fairly sparse (but maybe that shows a lack of interest in itself), but also those with large trees often don't respond to resolve differences. Having said that, I have had some good responses from both public and private tree owners (though rarely from those with no trees, perhaps not surprisingly!). I've also found that since I've had a public tree attached to my DNA results, no-one has initiated contact with me. I was hoping it would make it easier for people to find me. Maybe they do find me, but don't see the need to make contact. I would hope if someone found an error, they would let me know!

 

There's more than a degree of truth about having no need to make contact with Public Trees (those of substance of course) as things can often be assessed without the need for contact. However I agree I would hope someone would make contact if they discovered a significant error, just as I did last week.

To be fair I had communicated previously with the Tree owner (but a year or so ago) but after returning to view his Tree through a potential ancestor hint, I noted that he only had one of two brothers sentenced to transportation to VDL, the other remaining in the UK, married off and eventually dying there. I messaged to tell him I knew absolutely that both brothers had been transported, albeit 4 years apart. They never returned and gave chapter and verse on both. (I had even written story about them for the Midland Ancestor). Yesterday he got back to me and thanked me for the information, saying he had never been happy with the information assigned to the brother he had assumed remained in the UK. He asked for more information which I passed on earlier today.

I would of course hope that someone would do the same if they spotted something similar in my own Tree and as much as we often think we have things correctly researched, no one is immune from taking a wrong turn.

 

Oh, yes, I reversed the words at the beginning. It should have read "is it possible?"

 

I have found a similar (lack of) response to my public pedigree chart of just my direct ancestors. There was a single contact from a professional genealogist on the same day that I uploaded the tree but since then all contact initiation has been from me. The professional was pleased to get my response even though it has not resulted in any relationship confirmation and he did admit that about 99% of contact requests go unanswered. I have found that my experience is probably more like 50% but some of those responses may take a month or more before they arrive.

That, and so many testees (?) without any tree or only a couple of individuals included, makes me wonder why they submitted a sample in the first place. I have just received a tardy response from a potential 3rd/4th cousin (whose son shares 25% more DNA with me !?!?! see **) who is searching for a shared DNA match rather than looking for a MRCA within our trees. I am tempted to send a referral for LC but doubt that the masterclass would be read even then.

** Might that be because the potential cousin's spouse is also related to me or is it likely to be a simple variability of DNA inheritance?

 

Not being John McEnroe: more likely the former option as the latter would require DNA to bypass a generation

Presumably, the son has an additional segment or more?

Edit: I would think 25% too large for random recombination.

Phil

 

I first set out on both my own and my wife's DNA test results to associate them with Public Trees of Direct Ancestors, extracted from my main Trees. I found the after-math experience somewhat under whelming and wondered if things would improve by changing the associations to my main Trees, so that is what I did. OK I lost Thru Lines for a while but when reinstated things improved for the better, and whilst there are still the infuriating 'No Trees' or those with 2 or 3 persons showing (and even worse Private Trees) overall I feel much happier since taking this action. So -to coin a phrase ...'suits me sir' !

 

Parent has 35cM/2 and son has 43cM/3.

 

To the first I wanted to get away from just direct ancestors even though that was only playing a hunch, on a sort of 'the more the merrier' basis. I sensed that it would broaden the search so to speak (academic argument this would not stand up played no part). So my answer to your second point is my realisation that things did improve, with Thru lines especially (and yes I know they are based on Tree hints), and most noticeably with branch line connections and an increase in 'Potential' ancestry, all to be checked out.

Notification of matches also improved to a greater extent - whether these would have happened regardless of changing my Tree associations I do not know, but I sense not. I am aware of cm/segment strengths, but mainly limit myself to checking on 4th or better cousin listings - with the odd dabble to 5th. All this combined with conventional research is more than enough to keep me occupied and I leave the DNA techno speak to others, whilst in no way belittling their efforts.

 

When comparing my DNA matches with those of my son, I have quite a few instances where my son apparently shares more DNA with the person than I do. Mostly it's just a couple of cM, which I put down to possible missed bases or phasing algorithm accuracy which Peter has mentioned previously, but in several cases the difference is quite stark, e.g. one person shares only 9 cM/1 segment with me but 29 cM/2 segments with my son; two others share 10 cM with me, 25 cM with my son; another 24 cM with me, 31 cM with my son. On checking, I found that more than 10% of my son's matches in the range 21 - 35 cM show this sort of difference. The 'extra' DNA is not inherited from his father. I wonder if I can deduce anything about these matches, most of whom have no trees (or minimal trees) so I've no idea if we are related or not.

 

When children share more DNA and or segments than the parent with a relative and it can be demonstrated that the 'extra' DNA is not from the other parent, I suggest there are two possible explanations. The first is the additional cMs can in part at least be attributed to 'false positive' segments. In his paper on Small Segments Blaine Bettinger quotes "...researchers found that more than 67% of phased DNA segments shorter than 4 cM are false-positive segments! At least 60% of 4cM phased DNA segments were false-positive, and at least 33% of 5 cM phased DNA segments were false-positive." So although 'phasing' is part of Ancestry's algorithm it cannot be expected to eliminate all 'false positive' segments.

The second possible explanation is Ancesty's 'Timber' algorithm that strips out regions/segments that it considers to be over represented in the general population. How Ancestry applies 'Timber' to each person's DNA results is unknown but it would be done independently on each occasion meaning there is unlikely to be any connection/correlation between the parent and child on the application of the ‘Timber’ algorithm by Ancestry. One of the effects of the ‘Timber’ stripping is that whole segments become divided which can give the impression that additional segments are being shared.

Sometimes the additional DNA is a mystery like my 1st cousin and I triangulating 9cM with a match on FTDNA but Ancestry show me sharing 20cM with the same person! My conclusion has been to only take notice of the DNA segments actually shared between a parent and child and simply ignore the extra ‘mystery’ bits. Like Peter has suggests viewing the results on GEDmatch is really the only way to see the true picture.

 

What exactly do the segments indicate? My sister's daughter and I share 1762cM across 77 segments. The only other close family person to test with Ancestry is my male 1st cousin's daughter. We share 372cM across 14 segments. Two third cousins with the same 2x great-grandfather have 122/8 and 104/9. They are all on my maternal side so I have more information about all sets of grandparents. The next closest match after my niece and cousin shares 198/9 but I have yet to learn where he/she fits as that person has not replied to my message and the few shared matches are not helpful.