DNA Integration Superthread

I reread your comment, Private User Are you asking Geni to make a MRCA suggestion?
That would be interesting indeed...
Ancestry and 23&me and MyHeritage do it. FtDNA does not, to my knowledge.
I 100% believe they have more ability to connect us/answer the relationship path questions than what they do now.
I have several ancestors whose SNPs are known. If you match me there, I know who we share...

Cynthia Curtis, A183502, US7875087 MyHeritage has Theory of Relativity, Ancestry has Thrulines
-- their Common Ancestor suggestions are based on those - ie-- they are each based on the Trees provided, not the actual DNA.
FTDNA lets you link to Geni, where you can see if the Trees on Geni have any Common Ancestors - so yes, for those of us who have linked to FTDNA, we have that function.

So far, MyHeritage has not provided me with a single Theory. Not for a single Profile. Ancestry has.

Private User
Thrulines is based on dna matches and crappy trees.
MyHeritage at least uses Geni as a tree source so it is much more correct.
Yes, some of these companies have a known SNP for a common ancestor but they aren’t giving us a chrome browser and always have a Hire a Pro button at the top of the screen lol

Lois, I just looked at your pedigree here. It may be that not enough people have tested who are closely related to you.
Keep working on your tree and adding sources and work to use DNA to prove your close cousin relationships.

Cynthia Curtis, A183502, US7875087

Re: "Yes, some of these companies have a known SNP for a common ancestor but they aren’t giving us a chrome browser and always have a Hire a Pro button at the top of the screen lol"
MyHeritage gives us a chromosome browser, so are you only talking about Ancestry here?

In any case, where are you seeing any statement they "have a known SNP for a common ancestor" or what is your statement based on?

What I see on Ancestry is https://support.ancestry.com/s/article/Common-Ancestor-Hints?langua... includes no suggestion whatsoever that they are looking at any common SNPs. They insist they are just looking at the Trees.
"Here's how it works. If your great-grandmother is in your tree, and your match's mother of the same name is in her linked tree, we start comparing other details: their birthdays, locations, parents, spouses, and children. If they match up, your great-grandmother will appear to both of you as a common ancestor.

Because common ancestors depend on information from family trees, they're only as accurate as the trees themselves. To see common ancestors, you need to have a public family tree linked to your DNA test. For best results, build your tree back as far as you can."

Private User I literally have DNA matches with a MRCA showing for folks who have NO TREE.
I am not a machine but I know if you match me on a specific segment of copy 1 of chrome 8, you have Honeycutt…
The trees are crap. But the dna belongs to a specific individual.

I suppose I should say “I suspect” or IMHO
:)

Cynthia Curtis, A183502, US7875087
The fact that you can use DNA with a Chromosome Browser and think, we match on such-and-such segment, therefore they must be on such-and-such line
is totally different from any MRCA suggestions made by Ancestry or MyHeritage or 23&Me being based on that segment. They just look at Trees for any MRCA suggestions.

Private User we’re off topic

What can be done by humans with tools is different from what is being done automatcially.
Thrulines is suggestions but a MRCA with someone with no tree is based on what?
It’s based on dna.
But the key is triangulating meaning matching many on exactly the same segment and that cannot be compared without a chromosome browser.
Again, this is off topic.
You can pm me if you would like.

23&me has a very effective “tree”.
It is based on shared DNA and predicted relationships.
Position rather than shared names.
:)
And its a DNA site. NOT a genealogy site.

I think understanding that a suggestion a Site makes for a Common Ancestor is just based on Trees, not DNA, is on topic for "DNA Integration Superthread"

Also understanding that the Path that Geni shows to a DNA match has nothing to do with how the DNA connects the two.

These are two common instances of folks thinking a site is using DNA for something it is not actually using the DNA for.

The relevance of the first to the post by Private User ( https://www.geni.com/discussions/157595?msg=1628411 ) is that - FTDNA is not looking at the Trees,
And as I understand it, Geni is not looking at the raw atDNA data at all, but is just using FTDNA's match info.
Jan - I do not understand what you mean by "coalescing measures," nor what you are suggesting by " I suggest something small, maybe going back 4 generations only as starter. "
Also do not understand why doing that would mean "Then these emails would not really matter." - tho possibly would if I understood the other.

As for 23&Me -
I have now found https://customercare.23andme.com/hc/en-us/articles/360036068393-The... which does say
"If you choose to opt in to DNA Relatives, your Family Tree shows your close family members (3rd cousin or closer) that have been genotyped by 23andMe and who have also opted into the DNA Relatives tool. Additionally, any close family members with whom you have established a sharing connection will also be added to your tree.

Because the Family Tree predictions use only DNA and age, the identities of your predicted ancestors are unknown. When you determine which of your family members or ancestors links you to a particular DNA relative, you can manually write them in to the correct position in the tree. "

It refers to it as a new feature - but do not know when the article was written, so no idea how new it is.
Will check it out. Thank you.

"These are two common instances of folks thinking a site is using DNA for something it is not actually using the DNA for."
I 100% agree with that statement.
Having a path at Geni does not mean someone is really blood related.
This is probably why people freak out when their path is disproven

AH - this is what 2e&Me tells me now in info on Tree -
"To kick-start your tree, 23andMe looks at your closest 20 DNA Relatives and predicts their relationships to each other. After this initial tree prediction, new close relatives who opt in to DNA Relatives (2nd cousins and closer) will show up in your unplaced relatives list for you to add to your tree. From your unplaced relatives list, select the DNA Relative you wish to place on your tree, then select the empty placeholder circle where you wish to place them, or first add the correct empty placeholder circle if it does not yet exist."

After the initial bif, you apparently have to do your own placing, 23&Me will not https://customercare.23andme.com/hc/en-us/articles/360045637854?_gl.....
"If a close relative opts in to DNA Relatives after your tree was first calculated, they will be added automatically to your unplaced relatives list, from where you can add them to the appropriate position in your tree."

It says "You can find your unplaced relatives list by selecting the "Add a relative" button on your tree. You can move unplaced relatives in the list back onto your tree." - BUT clicking there just got me the statement I had not opted in to research. Which is correct, and I do not plan to do.

I have run out of positions to ADD A RELATIVE.

As stated before, I try to use all the tools available at each company.
Thrulines is a joke but offers clues.
My biggest issue, I think, is all the hints they give and there is no way to reject a Thrulines hint. Then people upload that bad tree junk here and other places.
I uploaded a family group once to Wikitree that way because I wanted the badge lol and that little Gedcom I made just for this particular person brought in all kinds of errors I had no idea were in my tree...
Ancestry used to give DNA circles which were confusing but it did batch people together.
Thrulines uses those crappy trees and is no way to reject a profile hint. I rarely use it but am really frustrated with all the same named folks that we have worked so hard to disambiguate and untangle only to see them messing up the Thrulines...
One example is Moses Swaim and his cousin Moses Swaim.
Another is Elizabeth Vickery and her sister Charity who both married Swaim... so I have specific examples.
But I am a loyal customer and it isn't for the Thrulines lol

Private User you said:
"They just look at Trees for any MRCA suggestions."

That is Ancestry for their Thrulines. Yes. And some of the positions they suggest are not even in the DNA matcher's tree so is based only on a rotten tree of someone you aren't even showing as a DNA match.

With MyHeritage and their Theory of Relativity, a MRCA can be proven because you can triangulate.
If you match one person with others on the same segment, that is triangulating and it only takes three people.
Matching MANY MANY people on that same exact segment is HOW we know where to look in the tree.
People married folks that were in the same location and rarely traveled in certain timeframes...

I know a segment of my chrome 5 maternal copy represents George Pratt.
Same for others.
I cannot find a parent for him because he is not in the records --- so I suspect a name change and have to work on the DNA until he is literally found in a historical record. I spoke with a pro about him. I am hoping to physically go where he lived and look for proof of him since nothing about him is digitalized apparently.
This is why we use BOTH genetic and traditional genealogy.
For anyone to say one is preferable to the other is just revealing something about themselves--
basically saying they trust and not the other but fact is we have to use ALL tools available for some lines.
Some will never be solved... Too many destroyed records, lack of access to records, incorrectly recorded events with spelling errors, etc.
DNA can prove a relationship but you have to build a tree to prove the relationship which is why matches will have relationship suggestions such as "half sibling, uncle, or nephew" but placing them correctly in the tee will give the accurate relationship for the match.
FtDNA has a matching and in-common or not in common feature for comparing DNA matches but I have to say I am most appreciative of MyHeritage's platform for DNA comparing.

We use the tools each company provides because each one has something to offer.
I speak from my own experiences.

My suggestion is to make your tree public with only the first few generations private so no one mismerges your parents. And only trust an ancestor is correct after you have triangulated.
There are DNA groups and genealogy groups on Facebook where folks offer assistance and explanations. That might be a good place for learning how to use your DNA matches to confirm your tree as well.
Moving on! I have some tree building to do :)

*digitized lol

There is a big difference between
MyHeritage uses the presence of triangulation in creating their Theory of Relativity suggestions
VERSUS
you can use triangulation you find to verify their suggestions

The fact that the second is true does not mean the first is true.
Everything I have seen and heard from MyHeritage says they just use the Trees to determine their Theory of Relativity suggestions.
Nothing you say above addresses that.

So yes, according to what they say, for both MyHeritage and Ancestry,
"They just look at Trees for any MRCA suggestions."

No.
You are not understanding triangulation, perhaps?

It is not my "job" to say how a company works.
But you have to UNDERSTAND TRIANGULATION to use DNA for confirming a tree...

You don't pay a DNA company to fill in your tree :)

Thrulines is NOT an example of triangulation.
Neither is Theory of Relativity.

According to everything I have seen and heard - and I have watched a number of Presentations by both MyHeritage and Ancestry and read a good bit of their printed information plus that of others,
for both MyHeritage and Ancestry,
"They just look at Trees for any MRCA suggestions."

Cynthia - If you have seen any place where MyHeritage says it uses DNA for their Theory of Relativity suggestions, please provide the link.

Anybody else who has seen such a place, also please present it.

--

Cynthia says above, in https://www.geni.com/discussions/157595?msg=1629215
"It is not my "job" to say how a company works."
BUT that is exactly what I am talking about when I say
For both Ancestry and MyHeritage, they just look at Trees for any MRCA suggestions.
And when she says that is not true for MyHeritage, then she is making a statement about how MyHeritage works.
Anybody who can explain that to her -- or can explain to me how I am misinterpreting what she is saying -- please jump in.

Thrulines proves nothing :)
Chromosome mapping is important
Tree building is important.
Triangulating is important.
You have to have three or more cousins to triangulate.
I have been at this for a long time but not as long as some.
we learn by watching, listening, reading, doing.
I am at the doing.
I am not here to teach anyone anything but to offer encouragement and share my experience and I ask for help when I get stumped.
Maybe that is the most important part-- to know what to ask and of whom.

Some comments on Thru Lines:

(1) It's only in operation for five generations or less.
(2) It is highly dependent on the quality of other peoples' trees. A bad tree and you get back nonsense.
(3) There are no DNA tools, like a chromosome browser, available.

So, it's better than nothing - for the novice. But I would be very saddened if FamilyTreeDNA decided that novices were their primary market and cut off access to actual information we are paying for as customers.

Detailed DNA analysis is not trivial to understand or perform. It cannot be "shortcutted". Even when a human being is doing the work, it requires a great deal of supporting genealogy effort to make it pay off. Even then, errors can occur, because sometimes the trees constructed for DNA analysis will be wrong. Later DNA data may force major revisions, especially when people are related to each other in multiple ways.

OK, I could not follow (with my limited experience and knowledge) everything that was said.

But what I meant with coalescence was basically that over many generations in limited "spaces"/locality, people become distant cousins naturally. Although I have studied this through algorithms, rather than through people or genealogical skills, this can be applied on Geni. From a statistical view, where I am more comfortable making this statement: atDNA matches becomes kinda meaningless without being able to pinpoint exactly what region (i.e. expected atDNA - which is always a chance working from now to before) and it would TRULY benefit all genealogy. I do not believe this should be accomplished by triangulation alone (but it is a cornerstone), rather statistically and programmatically (for me always = automatically), it should be easy to apply (with the end goal: if you can do 4... you can do 4 more, etc... data that is only growing, is not an issue right?)

It would need to introduce a further measure called probability certainty.

When you have a tight community of families intermarrying over time, I would think one would have to eliminate those smaller matches.
I have been using MyHeritage's tools more lately and their autoclusters seem to be similar to Ancestry's now defunct DNA Circles.
I have a particular surname I am having difficulty with pinpointing because there are so many testers/matches and one that is difficult because there are so few testers... with those, I NEED the smaller cM matches.
I don't really feel like I can go further in my tree with DNA matches at Ancestry because there just isn't enough information.
I have not been able to use WATO
Yes, I agree with your comment: atDNA matches becomes kinda meaningless without being able to pinpoint exactly what region (i.e. expected atDNA - which is always a chance working from now to before) and it would TRULY benefit all genealogy. I do not believe this should be accomplished by triangulation alone (but it is a cornerstone)
The rest of it I don't entirely understand.
I have several cousins here who have attached their DNA and when I look at their profile, it says "This relationship path has been confirmed by DNA testing. View details"
To me, these are golden.

I had not even noticed that --
"This relationship path has been confirmed by DNA testing. View details" ?
is at the bottom of the expanded path - so if you do not expand the path you do not see it

Private User -
"this can be applied on Geni." - on FTDNA, one can see details on a Chromosome Browser.
One cannot do that on Geni.- I am not sure Geni itself even has access to any of the Data that would be needed - So - how can that be applied on Geni?
(I always thought it was because it did not have such access that Geni's algorithm for suggested relatiinship only took total shared and longest into account, and so was so much poorer than FTDNA-s)