Please clarify criteria for calling a zero GD match

[jadebe]

For full sequence matches that FTDNA designates as GD=0, does that mean a match at every np, or just that they all meet the current phylotree build 17 sequence motif for the haplotype (U5a1a1d1 in this case)?

I have an extra coding region mutation and six matches called as zero GD. Seems a tad implausible that seven of us who've done the FTDNA full mt sequence test would all share an extra coding region mutation not found in published (GenBank) U5a1a1d1 sequences from multiple sources.

TIA

[chr]

Two sequences that match with GD=0 should be fully identical for every position, except 309 and 315 which are exluded from the comparison - see Learning Center https://www.familytreedna.com/learn/...-matches-page/ -> "mtDNA – Matches Page Questions" -> "On the mtDNA - Matches page, are only exact matches shown?".

Did you look in mitomap.org (https://www.mitomap.org/foswiki/bin/...n/SearchAllele) for your mutation?

[jadebe]

Two sequences that match with GD=0 should be fully identical for every position, except 309 and 315 which are exluded from the comparison - see Learning Center https://www.familytreedna.com/learn/...-matches-page/ -> "mtDNA – Matches Page Questions" -> "On the mtDNA - Matches page, are only exact matches shown?".

Thanks very much for your reply and for the link (which I should have found earlier, mea culpa). It appears that the 'fully identical excepting 309 and 315 rule' applies to control region (HVR1 & HVR2) matching but not to full sequence matching.

"For those who have tested the mtDNA Full Sequence (mtFullSequence), three differences are allowed. These differences include cases of heteroplasmy. Two high frequency insertion/deletion locations are completely excluded from difference counts. These are mutations at positions 309 and 315."

Did you look in mitomap.org (https://www.mitomap.org/foswiki/bin/...n/SearchAllele) for your mutation?

Yes, my extra control region mutation recurs at low frequency in African, Asian and Eurasian lineages, reaching greatest frequency (4.46%) in group W wherein it defines a tertiary classification.

Obviously, it recurs within U5a1a1d1 as well. While not found in any of the full mt GenBank U5a1a1d1 sequences, only one of those sequences does not have one or more other extra coding region and/or stable (non-'hotspot') control region mutations. Some of these mutations are shared by two or more sequences. Clearly, there is significant diversity/structure within U5a1a1d1 that won't be reflected in phylotree until the next update but is still relevant to estimating date range of mito MRCAs.

For genealogical purposes, for a client to have as accurate an estimate as possible in terms of mito MRCA timeframe (and make informed decisions regarding further research), it is necessary to know genetic distance. Allowing three differences to = zero GD, especially when those differences are in the coding region, is contradictory to that purpose.

Even when privacy concerns are taken into consideration, I can't see any reason why the full sequence matching model is allowed to call a GD of up to 3 as a zero.

[chr]

Thanks very much for your reply and for the link (which I should have found earlier, mea culpa). It appears that the 'fully identical excepting 309 and 315 rule' applies to control region (HVR1 & HVR2) matching but not to full sequence matching.

I will quote the full text of the FAQ entry and comment it, because it can be misunderstood easily, and probably you have.

On the mtDNA - Matches page, are only exact matches shown?
There are currently two sets of rules for mtDNA matching.

With "Filter Matches/Regions" set to "HVR1" or "HVR1, HVR2" the first set of rules ("Set A") is used, and as a result you get only exact matches - therefore no GD is being shown. Only Control Region is used for the comparison.

With Regions set to "HVR1, HVR2, Coding Regions" different rules ("Set B") are used. Not only exact matches are shown, but also matches with a small number of different mutations: a key value "Genetic Distance" is calculated from the differences, and only matches with GD=0, GD=1, GD=2 or GD=3 are shown to you.

For those who have tested HVR1 (mtDNA) or HVR1 and HVR2 (mtDNAPlus), only exact matches are shown. SmartMatching is also used to prevent a match between different haplogroups.

This is "Set A". Contrary to what the first sentence in the quote suggests, I understand it is also used if a kit is involved that has tested to Full Sequence level - as long as the Region-Filter is set to "HVR1" or "HVR1, HVR2", meaning you only compare in Control Region. They should better refer to the level of comparison and not the level of testing in this description. (It is pretty obvious that only kits can be compared at one of the three levels if they have that level tested at minimum)

For those who have tested the mtDNA Full Sequence (mtFullSequence), three differences are allowed. These differences include cases of heteroplasmy. Two high frequency insertion/deletion locations are completely excluded from difference counts. These are mutations at positions 309 and 315.

Tis is "Set B". Again, this really refers to the level of comparison, ("Filter Matches/Regions" = "HVR1, HVR2, Coding Regions") and not to the level of testing.

So, there is no "hidden" additional GD (as you fear), when comparing "Full Sequences" at Control Region level, because then also "Set A" is used. They just used some poor wording.

To me there are still questions open like:
Are 309 and 315 also excluded from comparison in "Set A"? (they don't mention it)
How is the GD calculated exactly in more complicated comparisons, like "1234.1A 1234.2Y" vs "1234d" (insertion, heteroplasmy, deletion) ... but I really stopped thinking about FTDNA GD some time ago.

[jadebe]

I will quote the full text of the FAQ entry and comment it, because it can be misunderstood easily, and probably you have.

I don't think so.

My mtDNA matching preference is set to 'Full Sequence Only'. That supposedly means I "will only see matches for the levels...tested" which was FMS. I can't set the filter to anything but 'HVR1, HVR2, Coding Regions'. Atttempts to pull up 'Set A' matches bring up the following message: "You have selected to hide results at this level. If you want to view these results, go to your Privacy & Sharing page and update your mtDNA Matching option."

If setting that pref to full seq only doesn't mean that all my supposedly zero GD matches shown are called from comparison only with other full mt seq's, then there is more than one FTDNA page that's in need of some major clarification.

If it does mean precisely that, then I must still conclude that there may be no true exact match between my sequence and any of the six shown as "zero" GD matches. In fact, there could be as many as 3 "differences" - in which case the likelihood of a common matrilineal ancestor within the last few thousand years declines precipitously.

Yes, it would be good to know how discriminating FTDNA's matching model is when it comes to the "differences" used to calculate GD. From years of experience working with mtDNA in another mammalian species, I'd certainly hope that they don't give equal weight to every "difference" but the FAQ gives little hope in that regard.

Thanks for the reply.

[chr]

If it does mean precisely that, then I must still conclude that there may be no true exact match between my sequence and any of the six shown as "zero" GD matches. In fact, there could be as many as 3 "differences" - in which case the likelihood of a common matrilineal ancestor within the last few thousand years declines precipitously.

I'm pretty convinced a Full Sequence match that has the "three allowed differences" will be shown as GD=3. It would be more clear if FTDNA would stick to one term in their Documentation and not use "Genetic Distance" and "allowed differences" in parallel.

About 309 and 315: seems to be not used in Control Region-only matching (https://www.familytreedna.com/learn/...evant-match/):

mtDNAPlus (HVR1 & HVR2) Exact Matches excluding positions 309 and 315
mtFull Sequence (HVR1, HVR2 & Coding Region) 0 through 3, including heteroplasmy, but excluding positions 309 and 315

[jadebe]

I'm pretty convinced a Full Sequence match that has the "three allowed differences" will be shown as GD=3.

I sincerely hope you're right about that. Will be contacting cust. svc. for clarification.

About 309 and 315: seems to be not used in Control Region-only matching

Indels in the stretch from ~303-315 are just too common for reliable match/mismatch calling, which is why they're among the several volatile loci excluded by phylotree from the sapiens tree reconstruction.

Thanks again