PDA

View Full Version : Improving FTDNA mtDNA matches and subclade assignments



GailT
01-19-2018, 04:16 PM
I have a list of topics that I hope to have FTDNA address in their mtDNA subclade assignments and mtDNA matching. I'm posting it here to see if others have comments or suggestions. Thanks.


1. Genetic distance for mtDNA Matches: Extremely common mutations and heteroplasmies are counted as steps in the GD. In some cases a common heteroplasmy is counted as two steps. Common mutations and heteroplasmies should be excluded when evaluating matches.

2. Some insertions/deletions are not being classified correctly in haplogroup assignments. This appears to be a bug in the software.

3. FTDNA uses Phylotree to define subclades, but Phylotree includes numerous paragroups defined by extremely common HVR mutations that revert frequently. This often results in people receiving a a subclade name that includes confusing and meaningless reversions, for example, U5b1-T16189C!-T16192C!. In some cases it results in people being assigned to the wrong subclade.

4. The ten year old table with estimates of the number of generations for most recent common maternal ancestor (MRCA) for mtDNA matches is not accurate for most people. The rate at which mtDNA mutations accumulate is highly variable, so there is no single estimate that is accurate for all people, but mtDNA matches can be useful if we do a better job of explaining how to use the information.

J1 DYS388=13
01-19-2018, 06:17 PM
I acknowledge that mtDNA is not supposed to be genealogically useful. But anything that would improve MRCA would be very welcome.

GailT
01-19-2018, 07:05 PM
I acknowledge that mtDNA is not supposed to be genealogically useful. But anything that would improve MRCA would be very welcome.


It is useful in some cases, especially if your maternal lineage has recent mutations. FTDNA need to do a much better job of explaining how mtDNA can be useful for genealogy.

J1 DYS388=13
01-20-2018, 10:25 AM
Another candidate for your list would be FTDNA's matching in the case of T16093C and T16093Y, which has always been very troublesome for me.

At least two experts have even said 16093 should not be used for matching at all.

And an academic paper has identified 16093! as a branching marker, which FTDNA does not recognize.