Should results be different in GEDMATCH one-to-one matches between FTDNA and Ancestry?

Question

I have tested with FTDNA and Ancestry DNA.  On GEDMATCH I have noted while doing one-to-one matches at 3 cm vs. default of 7 cm, that if FTDNA says it is not a match, Ancestry will have the Same result.  If FTDNA matches a kit, it is more than 50% that Ancestry will show a no match.  Where both match, usual the SNP count is higher and the start/end chromosome  range is broader at FTDNA and where Ancestry may identify one matching Chromosome, FTNA may show several more.  In one instance, FTDNA matched on four chromosomes and Ancestry matched, but on one different chromosome.

Comments

Hello Richard,

Are your results different while doing matches at 7 cM or higher?  

Are they different when higher than 12 cM?

Please read https://www.wikitree.com/g2g/199470/are-you-using-audna-segments-which-are-less-than-7-cm?show=245672

Thanks and sincerely, Peter

---

I only checked at 3 cm and 7 cm since 7cm was no matches, I would assume 12 would be the same.

Answer 1

That's why I have both an Ancestry kit and an FTDNA kit on GEDmatch.  While they do match, there are significant differences as you found, that make both useful.  It's clear that Ancestry and FTDNA don't just differ in which SNP's they choose to test, but in their philosophy of matching too.

It's also why I don't recommend uploading Ancestry DNA data to FTDNA.  I did that at first, and had limited matching.  Once I took the Family Finder test, and replaced my Ancestry uploaded data there, I had MANY more matches in the FTDNA database, which of course makes sense when you think about it.  But more than that, the kits behave a little differently in GEDmatch.  The FTDNA kit always finds more matching small segments, which can sometimes be useful, informative.

Comments

What I would really like to see is a Kit Merging Tool, that would merge all of the SNP's in each of your kits into one super kit.  This super kit would be optimal for comparing with any other kit, would be more likely to have the greatest number of SNP's in common with other kits, for the best comparisons.  I had hoped that GEDmatch was interested in doing that, but it seems to me they have been bogged down with Genesis and GDPR stuff.

A side effect of that is they could produce a report on which SNP's are in disagreement between the kits.  Promethease does that now, and reported 34 differences in SNP's they examined (I added both my Ancestry and FTDNA data).  I suspect they may have only checked SNP's in their database though, so there may be more.  I would imagine a merge of 2 kits would drop SNP's in conflict, but a merge of 3 kits might keep those SNP's with 2 out of 3 in agreement.  A side effect is that that might allow a 'quality' test, which companies have the most SNP values dropped or kept.

Answer 2

There are some minor differences if you upload kits from different sites to gedmatch. If you decide to have several kits on there please mark all but one as research kits. Otherwise you will frustrate your matches who see you as repeats and lose valuable places in those match list views which have a count cut-off.

Comments

Thank-you, I didn't see a block to check, but will look again

---

okay, found the block to check and now I have another question.  There is a red remark in this block: "We do NOT encourage the use of the 'Research' option with normal kit uploads, since it is not consistent with the free exchange of genealogy information." 

So I guess I should ask what is a "normal kit uploads"?  I know I see many people with multiple kits in my matches, so your comment was a bit of a surprise.  I have no intention to cause a system delay, cut-off, or any cuts-offs due to multiple kits, but believe the developers need to clarify where they believe this option should be used.  The implication of the red comment is they would prefer we not use this option.

---

They also have an FAQ in which they discourage duplicates, so then the question arises of what is considered a duplicate. I don't think gedmatch addresses the issue very clearly. However, if you take a look at your own one-to-many list I expect you will find several instances of people who are taking up several spaces on your list because they are showing kits from each vendor they have tested with. For each extra kit they post someone else drops off the bottom of your list. The topic comes up quite frequently on facebook discussions about gedmatch and, as with most discussions, there are differing views. There is no rule to say only show one kit, but just be aware of how they appear to your matches if you show them all.

---

okay, I'm convinced, done.