Suggesting Alert

Tagging Mike Stangel

We're aware of the situation and have provided FTDNA with the mechanism they need to update DNA, but they have chosen not to do so. Instructions for updating your results by unlinking and re-linking your account are here: https://help.geni.com/hc/en-us/articles/115005316387-How-do-I-updat...

Dear Mike,
Thanks for the post. My point is that given the situation and the fact that many GENI users are not aware of the circumstances, there should be a way of briefly explaining the problem and its workaround so that users can take a proactive response. Is there some systematic way for GENI to communicate about matters of import to the entire community?
Cordially,
Adam

Private User I think in a lot of cases users link their FTDNA account and never need to update it. I'm hesitant to blast everyone (even just the DNA-linkers) with instructions for something they may never need to do. Maybe we can find a way to work this into the text on your DNA tab....

Mike Stangel Thanks much for the attention. Many serious amateur dna researchers start with one test at FTDNA then upgrade incrementally (especially if you consider that the prices of the Y700 have been coming down and there are record numbers of Y700 adopters in 2019 and continuing into 2020). Your presumption is not what I have encountered in the course of administering an FTDNA project of 200+ persons. This discussion ties in well with several other problems with the way haplogroup data is functioning on GENI, which is less than ideal at present. In my line of research, haplogroup connections using advanced testing (which is far more accurate than the broad haplogroup categories we once used) are perhaps the single most important strategy offered by GENI for building deep trees (and there are many excellent GENI features so that is saying alot). Here are the other two aspects of haplogroup data needing improvement: 1) the system is not picking up haplogroup conflicts well enough (see this discussion for an example: Lazar Charlap), and 2) under the DNA tab, when looking for a list of people who could be tested for y- or mt-dna, the system should be able to show whether any persons on the list have already tested and connected their y- or mt-dna to GENI (sometimes these lists produce hundreds of hits and it would save enormous time to see instantly that someone on the list has already provided the needed test results to GENI, instead of having to inspect each profile on the list one at a time and then happen to see that the person has already been tested; this relates to the distance to which the haplogroup data propogates up the tree so if I am looking for a descendant of a person from the 16th Century, a haplogroup logo for that person does not appear on the profile even if there is a descendant who has tested). I am speaking from personal experience when I make these suggestions, which are not excessively complex to implement, and which would make working with GENI much more effective. I can understand not wanting to abuse the power of communication that you have with the entire community. Important improvements or changes to the system should be publicized in a periodic newsletter so that researchers who may have been frustrated by prior experience can know things are working better now. These suggestions are very important for GENI to consider so please feel free to contact me further for more details or if I have not made myself clear enough. Appreciatively, Adam

Mike - think your idea to include a comment in the DNA tab info about the need, if one gets new results, to disconnect and then reconnect if one wants those results included is a good one. Writing it out there and/or providing there that link you give above.

Private User thanks for your detailed explanation. We can definitely address the issue with the list view now showing who has already tested DNA.

The haplogroup issue is a little more complicated, because FTDNA only sends us STR markers, not SNPs. They'll tell us a haplogroup but they won't tell us whether that's assumed (based on a small number of STRs) or confirmed with SNPs. We know the STR haplotree well enough that I think we could know the difference, however what we don't know well is the SNP haplotree. For that reason, we do not use haplogroup for matching (or rejecting matches) even when the haplogroup is SNP-specific rather than assumed from STRs. For example in Y-DNA I believe CTS1006 and CTS10941 are both variants of M170/PF3715 but I don't know of a definitive resource for the entire SNP haplotree (we once tried scraping the ISOGG pages but quickly discovered there are too many inconsistencies even between pages on the ISOGG site to succeed). That means we wouldn't be able to tell you how closely related a CTS1006 test-take is, to a CTS10941 test-taker.

Feel free to correct any misunderstandings in what I've written above; I've had a fair amount of experience with this but I don't claim to be an expert.