Upload DNA data for more information
edited September 28, 2020 in Suggest an Idea
Aleksander Tsoi said: Hello, I have my DNA Data by Ancestry. It would be great to upload DNA Data to FamilySearch data and find additional information.
A van Helsdingen said: FamilySearch does not have any capability for DNA matching. Their focus is on traditional genealogy. Quite a bit of technical knowledge is required for a company or organization to provide a reliable DNA matching service.
There are other sites that will accept DNA data uploads. MyHeritage, until 1 December, is accepting DNA uploads for free. If you are a LDS member you can then link a family tree on MH to the FamilySearch FamilyTree.
Other sites accepting DNA uploads include GedMatch, FTDNA and LivingDNA.0
robertkehrer said: Aleksander,
A Van Helsdingen is correct, Familysearch does not have any DNA upload or services.
Out of curiosity, Can you outline for me what specific services or information you would anticipate from FamilySearch after uploading your DNA file? What user experiences are you looking for?
Alahärmä said: I would like to attach strips of reconstructed DNA onto my ancestors on the tree, with logical proofs derived from tests of living descendants of course. That would provide experimental evidence to support the document based family relationships.
I do not know of any service that would allow such reconstructions. Those would also help cooperation among different branches where the descendants do not prefer sharing their full DNA with all people in the internet, but would like to make their findings public.
Now that we do have many companies who offer the result data for downloading, there is no need for FamilySearch to become yet another one. However FamilyTree has the highest quality source referencing system in the world and such possibility of attaching DNA to the ancestors would be one step further.
Now that I think of it, one could attach such DNA data as memories in the ancestors' profiles. Is that allowable? That would be one simple step towards the vision. However that would not allow any automated comparisons or validations between people.
Think of Borland Genetics. https://isogg.org/wiki/Borland_Genetics0
robertkehrer said: Alahärmä,
It would be inappropriate to upload raw genetic data (SNP calls or full genome) as sources in the Family Tree. FamilySearch, nor the FamilyTree tool specifically, does not have systems to enable the upload, proper storage and handling according to widely established privacy and consent standards & laws, nor does it have tools to facilitate the analysis of raw genetic data.
Memories is also NOT an appropriate place to store raw genetic data. Again this data has serious privacy and consent implications and is governed by specific national laws that vary by country that Memories is not built to deal with. As you note, there are no genetic analysis tools in the Memories system. It is built for photos, stories, documents and other human consumable data types, not raw genetic data.
Adding proof statements that incorporate genetic evidence to ancestors in Family Tree is absolutely acceptable. As you are aware, genetic evidence can, if the shared quantity is sufficient, validate only the amount of shared DNA through a putative recent common ancestor. It cannot, by itself, unambiguously validate any relationship types, except for Self & Parent-Child. Relationship validation proof statements must include genealogical evidence to eliminate any ambiguity.
There are a number of genome reconstruction tools under development, including Kevin Borland's work. Amy Williams from Cornell University delivered a talk at RootsTech 2020 SLC demonstrating the excellent progress her team has made in genome reconstruction and validating the finding of cousins using a Grandmother's reconstructed DNA. Much of this is still in the research and development phase with only limited accessibility to avg users yet, but this technique is expected to be a powerful future tool.
Please do not upload raw genetic data, reconstructed or otherwise, onto the existing FamilySearch systems.
sr product manager, genetic genealogy0
Tom Huber said: I've noticed an interesting trend with DNA -- that while it can provide assistance, especially with Y-DNA and mtDNA results, in determining lineage, it is far from perfect.
For instance, My Y-DNA came up with an uncommon haplogroup. In working in a group on FamilyTreeDNA's site (not related to the tree management program) I saw that people in the group were not registered with the same haplogroup. The help came from adding some additional SNPs to the results, which then matched the haplogroup with others group. The 23 and Me results were also confusing. I ended up with the same uncommon haplogroup as originally identified by FamilyTreeDNA, but my son, with a test from 23 and Me ended up with a different Y-DNA haplogroup, yet his admixture showed he was indeed, my son.
In the early years of Dr. Gates' PBS series, DNA was relied heavily for chasing down ancestral homes. The current season of Finding Your Roots does not mention DNA, except in passing and does nothing for the persons, unless they had a connection to another of Gates' guests through DNA.
So at this point in time, I am extremely hesitant to record any DNA in FamilyTree (it can be recorded as a custom fact), even though common accepted belief is that both mtDNA and Y-DNA can go back generations with little, if any variation. What is my Y-DNA haplogroup should be the same for my direct paternal line, except the Y-DNA for my son and myself says otherwise.0
A van Helsdingen said: 23andMe's test is not a complete or accurate Y-DNA test. It is an autosomal test with some Y-DNA and mtDNA added on as an afterthought.
FTDNA is a reputable testing site for both Y-DNA and mtDNA. They test the entire mitochrondia and enough of the Y-DNA to make accurate conclusions.
Was your haplogroup I-M223? I seem to remember someone with your name or something very similiar participating in the FTDNA I-M223 group some time back.0
Tom Huber said: Yes. I'm still a member of that group. My Y-DNA haplogroup is now I-BY3760
after adding the SNP results to my Y-111 test results.0
terry blair said: Tom, the problem is not with you and your son, but rather with the testing and reporting protocols of the testing companies. If the haplogroup progression is thought of as a tree, I think you will be further down the tree than your son. This is because 23 and Me only tests and reports a few haplogroups and these are usually higher up in the tree. The way to deal with the difference is to: ignore the 23 and Me results; write out the longest path for either of you from the root of the tree and then see where the other person is along the path; or both of you get tested to the same level at the same company. This is a known and commonly questioned difference between 23 and Me and other testing companies. I0
Tom Huber said: By the way, I agree with your assessment of 23andMe's test (and thank you for bringing it up).
Down the road, I'm going to have my son's Y-DNA tested with FTDNA, just to see what the results are.0
A van Helsdingen said: My specific branch within I-M223 is I-Y4760.
Out of curiosity I had a look at the phylogenetic tree on FTDNA and at the websites YFull and ScaledInInnovation, and I see that our common male-line ancestor lived approximately 10,400 years ago and most probably in modern day Germany. Shows just how old haplgroup I is compared to other European haplogroups.0
Tom Huber said: Thanks for the information. Can you post links to those two websites? I appreciate it.0
Tom Huber said: Yeah, the major companies involved with testing do things differently when it come to analyzing the results. MyHeritage's results had me with sub-Saharan ancestry when none of the others (Ancestry, 23andMe, and FTDNA) gave any indication of that ancestry.0
A van Helsdingen said: So I had a look at the FTDNA haplotree and saw that the most recent common ancestor of our two branches was called I-CTS616.
YFull estimates 11,900 years ago for the formation of I-CTS616 and 10,400 years ago for the man who is common ancestor of all of it's subgroups. https://www.yfull.com/tree/I-CTS616/
You can also see your subclade's page on YFull (they use Y-BY3760 as an alternative name for it rather than its "main" name) : https://www.yfull.com/tree/I-Y43928/
ScaledInInnvoation takes data from FTDNA, YFull and scientific research to make maps and charts showing how different haplogroups moved over time. http://scaledinnovation.com/gg/snpTra... There are no separate URLs for each halogroup, you have to type in "I-CTS616" for instance to see results.
I hope this is interesting.0
A van Helsdingen said: But we are talking about autosomal DNA here, and presumably only a small percentage? As you may know anything under about 5% on an autosomal ethnicity test should be treated with great suspicion.0
Tom Huber said: Yup on both counts.0
Brett said: Alahärmä
Directly from a "Knowledge Article" in "FamilySearch":
Please note that FamilySearch does not endorse or recommend any commercial DNA applications for genetic genealogy. Nor does our Family Tree provide specific features to post or link DNA groups based on DNA matches.
Using DNA in family history research
Juli said: Just a nitpick: consumer-grade DNA tests cannot even really validate "Self" and "Parent-Child". DNA says my sister is me and that she's my child's mother.0
Juli said: If you bring MyHeritage's admixtures into it, you have to start ignoring percentages a lot greater than 5... (MH puts me at 12% Scandinavian. I have zero actual Scandinavian ancestors, going back to third great grandparents, a fact supported by the admixtures reported by both 23andMe and Ancestry. Those admixtures aren't perfect, either, but they're a lot less ridiculously wrong.)0
A van Helsdingen said: Any good (and even not so good/scams) DNA test can determine a sibling relationship without difficulty. Something is seriously wrong if they cannot.
You should match around 3500 cM with a full sibling, i.e 50% of your DNA.
If you match 100% (i.e around 7000 cM), then either you're identical twins or there's been a mix-up with tests.0
A van Helsdingen said: Over-estimation of Scandinavian heritage is a phenomenon going back several years. There used to be people being told they were majority Scandinavian, particularly on AncestryDNA, when they had no genuine, recent, heritage from Scandinavia.0
ATP said: "Adding proof statements that incorporate genetic evidence to ancestors in Family Tree is absolutely acceptable. As you are aware, genetic evidence can, if the shared quantity is sufficient, validate only the amount of shared DNA through a putative recent common ancestor. It cannot, by itself, unambiguously validate any relationship types, except for Self & Parent-Child. Relationship validation proof statements must include genealogical evidence to eliminate any ambiguity."
For any who may not have followed the recent very public example of how validation proof statements and DNA were used in identifying the skeletal remains of England's King Richard III by the University of Leicester, the following links may be of interest.
Richard III left no known descendants. The mitochondrial DNA of 2 descendants of his sister, Anne of York was used to verify the shared mitochondrial DNA of Richard and Anne, inherited from their mother, together with the documented paper trails of those 2 descendants of Anne of York to prove that the bones that had been discovered belonged to Richard III.
robertkehrer said: A van Helsdingen, Of course you are right about the tests, however a half sibling on the high end of shared DNA and a full sibling on the low end overlap in the amount of DNA they can share. Better to unite the DNA evidence with the genealogical information to reduce the ambiguity.0
Alahärmä said: Tom, wait for a €150 black friday discount offer from Dante Labs and order whole genome sequencing.0
Alahärmä said: Then use WGSExtract to derive output data files from the whole genome BAM file.0
Alahärmä said: Yes. The support for raw DNA data snippets could be completely agnostic in regards to any data provider. Such data is not commercial by nature.0
robertkehrer said: Under the laws of most countries the user owns their DNA and the testing company must obtain consent from the user to hold the DNA, to process it into a raw genetic data file and to analyze the data. Because the user owns their data, all the testing companies offer users the ability to download their DNA data and multiple 3rd party companies accept uploads of the most common file types.
Prior to accepting the upload of any genetic data file all companies build their systems to obtain and maintain proper consent from the user. They build their systems to meet or exceed industry standards practices around the secure holding of genetic data and the preservation of privacy...and they address a host of other concerns.
FamilySearch systems are not built for nor intended to accept the upload of genetic data, nor to obtain and maintain proper consent. Likewise they are not built to analyze this raw data. Because of that please do not upload genetic data to the current FamilySearch systems. You will need to conduct your genetic analyses on other services but it is completely appropriate to include information from the analytical output (ex. quantity of share DNA, estimated relationships, etc) in your proof statements in Family Tree.0
Juli said: AvanH, that's exactly my point. I have an identical twin. DNA cannot tell us apart. Therefore, DNA cannot, in fact, definitively prove even "self" or "parent-child" type relationships.0
David Newton said: It should also be pointed out that siblings and double first cousins also overlap in the amount of DNA that they can share.0
A van Helsdingen said: Identical twins are of course the exception/special case with DNA. From a DNA perspective you and your sister are the same person. The children of identical siblings are not genetically 1st cousins but rather half siblings. It does not invalidate DNA testing as a form of evidence, it just means the possibility of identical siblings must be considered by genealogists when they use DNA evidence, just as many historical records have their quirks and special cases. No one here is saying that DNA evidence is definitive and should replace traditional genealogy research.0
Tom Huber said: Indeed. Privacy Laws must be obeyed and therefore DNA of living persons should not be exposed except as a site is able to properly manage under the control of the person, the information.
What Robert is saying is that FamilySearch is not set up to do that.
In addition, the interpretation of DNA information (of which Y-DNA and mtDNA are the only useful reports) and determining ancestral homes and haplogroups, varies with the reporting company. In some cases, the reporting is highly qualified (as in the case of FTDNA) and in others, the reporting is not based upon enough of the information in the raw data (as in the case of 23andMe). There is a discussion above between me and several others that point to the problem.
Admixture reports can be useful, but only over a very limited number of generations. The dilution of the DNA is such that after three or four generations, the results are inconclusive and even over two or three generations, there can be problems using DNA for genealogical purposes.0