Allow DNA connection details to be entered into the private space.

LegacyUser

Loren Brownlee said: I use several DNA sites to identify DNA matches. This has been super helpful to verify researched family connections and to identify where to look for additional ancestors and cousins. Being able to record this in family tree would provide more analytical data to aid and direct research efforts. Currently this seems to be a shortcoming in the family tree.

LegacyUser

Tom Huber said: Welcome back to the community-powered feedback forum for FamilySearch. FamilySearch personnel read every discussion thread and may or may not respond as their time permits. We all share an active interest in using the resources of this site and as users, we have various levels of knowledge and experience and do our best to help each other with concerns, issues, and/or questions.

DNA connections depend upon the type of test that was run. For instance, an admixture (AKA autosomal) test (used by Ancestry and many testing sites) is only good for about three or so generations before the amount of usable DNA is degraded to the point of being almost useless. It is only really effective for two or three generations at the most.

mtDNA follows the mother's maternal ancestry. Y-DNA follows the father's paternal ancestry. Both can go back for generations and are parts of haplogroups. However, not all haplogroup discoveries are equal. 23andMe does not fully test for mtDNA and Y-DNA ancestry and the named Haplogroup can be misleading. FTDNA does a better job but is dependent upon the test used. Even using the Y111 Y-DNA test, I had to add several SNPs to get to the most accurate Haplogroup.

While it is useful to have this information, only a test of the actual person can be used for that person, who likely is currently living and therefore not something that can be seen by anyone but the person who entered the living person's profile.

At some future date, when DNA test has matured far beyond what it is now and companies are using the same interpretive information (the raw data is the same, but the interpretation varies widely), and somehow being able to fully determine ancestry that wasn't interrupted by adoption (which may or may not be known) through the paper trail... then, and only then could I see where the information would be beneficial.

LegacyUser

robertkehrer said: Loren,
Thank you for your request. As you are aware there is currently no way to properly save genetic information into the FamilySearch system. The unique nature of both raw genetic data, and the subsequent analytical data generated from it, will require appropriate privacy and security controls as well as unique interfaces in order for it to be used as evidence in forming conclusions in Family Tree.

I would be very interested in your thoughts around what specific data you wish to bring over to FamilySearch, how you envision that data presented and to whom it might be visible on the system. Any thoughts you can share would be valuable.

-Robert

LegacyUser

Loren Brownlee said: Hi Robert,

I see DNA information but not raw data being recorded by users in their private space. So the living relatives who have common DNA can be identified. Basic information might be something like a percentage of DNA or number of centimorgans in common. If multiple siblings have been tested then the sibling with the highest amount of DNA for common matches would be used.

I am an ancestry user and the key difference between ancestry and family search is that the tree in family search is one world tree whereas in ancestry its a collection of private trees, many of which are inaccurate. With family tree sharing contributors and having such an excellent database of records and sources so easily attached, it is becoming increasingly more accurate.

Identifying the MRCA for DNA matches is aided a little in ancestry with their thru-lines technology but its biggest issue is that each match's tree is incomplete, private or inaccurate. If family search could be used by individuals in their private space to be able to record common matches and work to connect them into the tree then MRCA could be more easily identified.

I understand that the private space is visible in the mobile family tree but not yet in the web version. Bringing in the possibility to add basic DNA from common matches would enable more analytics. For example if familysearch can see all my common matches identified in the tree and has at least partial data to identify a few generations back from each match then connecting the DNA lines can be possible or estimated. This could be a huge benefit to all.

Using the private space would ensure that DNA information would remain private and only visible to the creator of that information. Then family search would be able to suggest to me possible links to my common matches as disconnected partial trees get plugged into the main tree.

LegacyUser

robertkehrer said: Loren (and anyone who wants to chime in),
I want to make sure I am extracting accurate learnings from your comments Here's is what I'm hearing:
1) An Aggregation Site
It sounds like you are looking for a place where you can privately gather data from your test provider and conduct additional analytics on it. I observe that this could be profoundly powerful for those who have tested at multiple providers.

2) Power of FamilyTree
The collaborative tree has clearly been driving a completeness and accuracy unavailable on other private tree websites. It's ability to connect living cousins to each other and their common ancestor is accentuated in the genetic genealogy domain. Combining "basic" genetic data with Family Tree will allow additional analyses by users.

Questions:
A) In your view what specific data would you want to bring over to FamilySearch from your testing company? (Shared cM, Cousin info, Shared matches info, segment info (Chr #, start/stop positions, cM length, # SNPs?)...and tightly related to that...

What specific "more analysis" would you want to conduct on FamilySearch.

C) I'd like your thoughts on the private vs shared nature of things. Should anyone else be able to see any of the data you bring over to FamilySearch? How about once you use that data as evidence in a conclusion? Should any of that data be visible to others if a conclusion is based on it? What data?

-Robert

LegacyUser

Tom Huber said: The problem that currently, DNA testing is only done on living individuals.

There is a legal issue involved with providing information for any living persons and since FamilySearch is international in scope, the most stringent privacy laws must be observed.

While there is hope to be able to provide connections to living relatives, I suspect the logistics is complex, at least to the point where two people have to both agree that the other can see certain living person profiles. Without permission, I do not think the privacy laws would allow just anyone to observe another living person's profile.

So much for that legal issue -- that would have to be worked out somehow and I think that Loren's desire would include any person whose DNA came close to matching his. (He'll have to confirm.)

While my generation (I'm 75) will not have a lot to worry about, mostly because most of us will be deceased in the next 25 years or so, at that point, the material on my profile becomes public with my death.

But for the most part, those of later generations than mine are under the laws that limit who can see this information, especially when it comes to genetically-based health concerns..

Private health issues are exposed in the genetic code -- especially the raw data. 23andMe deals with the health issues on a regular basis and while individuals are protected (no names are attached to the studies), exposing the raw data in Family Tree could be very problematical. United States' laws alone carefully guard against exposing one's health data to the public.

Then there is the issue that no one company interprets the data the same. Even the autosomal admixture results are all over the board. One company said I had sub-sarahan ancestry, which is completely unsupported by the paper trail and every other company provided no such connections.

About the only thing that can be trusted is the raw data, because the interpretations are not consistent between companies.

mtDNA and Y-DNA are more consistent, but those companies (like 23andMe) that provides Haplogroup information as a side feature do not test deep enough to make their claims completely accurate. As I mentioned in my first reply, even FTDNA's Y-111 test was insufficient and I had to have additional SNP's included to get to the point where my Y-DNA Haplogroup matched those in my group. My son and I both took the 23andMe test, but his Y-DNA Haplogroup was nowhere near mine, providing solid evidence that their Haplogroup identification was no accurate. Comparing our two charts clearly showed our genetic connection, and therefore, the Haplogroup should have been the same that 23andMe provided to both of us.

LegacyUser

robertkehrer said: Tom,
Thanks for your comments. I think there is value in being very clear we are NOT talking about anyone bringing raw genetic data files over to the FamilySearch website in this discussion. You are absolutely correct about the privacy risks, particularly to health data, inherent in these files.

You also note a living to living sharing model providing shared connections. You rightly note that such a thing does not exist currently, although a model and precedent has been set with the Relatives Around Me opt-in model implemented in the mobile clients. In any case, anticipating any such Living-to-Livign model may also be outside the scope of this thread. He was clear with his title that this was a discussion about his private space.

You also accurately point out that admixture, also called ethnicities, calculations are just estimates and vary across companies. However, I believe that the these are of much lower interest and utility (not completely irrelevant) to genealogical research. mt and Y results are much more useful in specific workflows and should probably be discussed as conclusionary evidence here. The core of what I think Loren is asking for, since he mentions "DNA matches" "family connections" and "cousins", is the analysis of shared segment data between cousins.

As I interpret what Loren has asked for, it was a private place to gather his cousin match information from one or more testing companies, to tie that data to the family tree (meaning to deceased PIDs and living cousin PIDs in his private space) and to gain the benefits of Family Tree and FS historical records in organizing and making conclusions visible to him only.

In my return question, I did broach the topic of whether any of of the data he would anticipate bringing over and using in a conclusionary way should ever be visible to public and what that might be.

Also please note, this is purely exploratory so I can learn what users see and expect...not any kind of product promise here.

LegacyUser

Loren Brownlee said: Thanks Robert for your comments. You are correct in what I have tried to communicate.
A) I would only want to enter basic DNA match details for my common matches. So for example when I have a group of common matches, people who share the same DNA pieces as I do, then the implication for me is that we share a common ancestor. I would want to mark them similarly as a group. It would be helpful to record for each, the percentage of common DNA or the centimorgans in common and an estimate for the relationship like say 4th cousin. The group indicator would be helpful in a visualization where it would be nice to see several common matches and their trees to look for overlaps or locations in common where I could direct my research efforts to find our common links.

so as I mentioned, the visualization would be helpful to direct my research efforts. Also since familysearch can see my private space, they could aggregate common match data from various patrons like myself and make hints of where potential links might be to other common matches in an effort to suggest MRCAs for your patrons. Once this is done, duplicates could be merged and cleaned up and the overall tree integrity improved. A visualization could be for example a series of fan charts for my common matches to assist with eyeballing where our trees might meet or overlap.

C) the private space keeps things unique to myself so avoids privacy issues. So no-one should be able to see my data in that space including DNA information. If I should die it would be nice if that private space could be inherited by another patron or relative. As my living people die and move from my private space to the public space then I could create dummy records in my private space to continue with that DNA information if the common match is still not resolved. Once a DNA match has been identified in the main familytree and the MRCA is known then I consider the match resolved. Even once my match is resolved, tracking this as a common match will help as other common matches become known to me and I will know where to look for connections. Through the use of dummy records in my private space I can still track deceased DNA relatives.

When making relationship connections in familytree we can already include reasons and notes so we can mention DNA evidence. I do not think we can reveal any DNA information publicly. The tree itself suggests DNA connections and that is sufficient for public consumption. We also need to consider that DNA testers such as ancestry 23andme etc have rights over that data too.

LegacyUser

Adrian Bruce said: Robert - my major worry with all this is that there is a huge amount of "stuff" (to use a technical term!) associated with DNA. If FS is going to start going down that route, it will have to learn a huge amount - for what? To duplicate what Ancestry and a few other sites have done. And that will need the transfer of the raw data (much like GEDMatch does). But below you seem to be (sensibly in my view) excluding that.

So I'm not wholly clear what anyone would be left with other than, basically, notes dedicated to DNA with all the work left to the user.

Guess I'm not seeing the potential usages in any detail.

LegacyUser

robertkehrer said: Adrian,

As you surmise, uploading the raw genetic data files like GedMatch, is outside the scope of this inquiry (by either myself or Loren). There are a number of very useful tools that allow users to bring over pieces of the analytical output (ex shared cM or segment data) and use this to help the user come to solid genealogical conclusions. I'm thinking of tools like the Shared Centimorgan Project, DNAPainter's Chromosome Browser, the very useful What Are The Odds tool, DNAGedcom, Genome Mate Pro, and a host of others.

It turns out that just seeing some shared DNA and concluding that you are related is where most hobbyists stop. There are some very good efforts underway in the industry to provide tools that help the hobbiest go one step further into genealogical accuracy and actually triangulate segments, or statistically estimate most probable relationships. I think these might be the kinds of additional analyses Loren is seeking to do. He also mentioned the power of forming these genealogical conclusions in the context of FamilyTree ... and I would add maybe with aggregated results from multiple testing companies.

I think these are the kinds of things I'd like to get user feedback on.

-Robert

LegacyUser

Adrian Bruce said: OK - that sounds like some useful ideas for usages.