Monday, November 27, 2006

FTDNATiP™ Interpretation

These reports perform calculations for the Most Recent Common Ancestor (MRCA) of two people, using mutation rates for individual markers (known only to FTDNA).   FTDNATiP™ reports show actual probabilities based on which specific markers you mismatch with another participant.   A person who has a one-step difference on a slow marker would get a different report than a person who has a one-step difference on a fast marker.

It might seem intuitive that for a given marker distance, that distance on a fast marker would suggest a more recent TMRCA than that distance on slow moving markers.   That would be wrong.

Ann Turner has addressed what those different results might be depending on whether your marker differences occurred on slow markers or on fast markers in a couple of recent GENEALOGY-DNA List posts and has agreed that I may quote from them.

"With the mutation rates published by Sorenson and their TMRCA calculations using individual marker rates, you can see how little difference the mutation rate makes.   Plus the difference is the opposite of what people sometimes expect.

For example, I checked the TMRCA for a 12-marker test (called 11 markers at Sorenson) when everything matched except DYS388 (rate = .000379) or DYS390 (rate = .004181), about a 10-fold difference.
Prob          388          390
25% 18 20 generations
50% 32 35
75% off the chart
So a 10-fold difference in the mutation rate only changed the median by 3 generations, a small amount compared to the broad uncertainty range.

For a 35-marker match (38 loci, counting multi-copy markers), I used DYS426 (.000271) and DYS449 (.006464). Here the difference in mutation rates is 24x, but it's averaged in with many more markers, so there's less of an impact.   The numbers are rounded to the nearest generation -- there would probably be a *slight* difference if fractional generations were given, but the emphasis is on the word slight.
Prob          426          449
25% 6 7 generations
50% 11 11
75% 17 17
The mutation rates used by SMGF are in the poster Direct determination of mutation characteristics of Y chromosome STR loci".

You may examine that poster by clicking this link:
Direct determination of mutation characteristics of Y chromosome STR loci

She later considered whether two fast markers differing in a haplotype would yield a markedly different result than two slow moving markers -

"Changing two slow markers (DYS388 and DYS426), I get 12, 17, and 25 generations for 25%, 50%, and 75% cumulative probability of finding an ancestor.

Changing two fast markers (DYS449 and DYS458), I get 13, 19, and 27 (longer but not dramatically so).

The most important factor is the overall mutation rate, since we must take into account what has happened with the marker(s) that changed AND all the markers that did NOT change.   Thus, if two people have a mismatch on a "slow" marker, the MRCA is actually more recent.   This is hard for me to wrap my head around, but one way of explaining it is that a lot of "fast" markers did NOT change, and we would expect to see more mutations on those markers if the ancestor were in the distant past."

Her conclusion:   "Bottom line -- don't worry about whether the marker has a fast/slow mutation rate, just consider the total number of mutations."


Post a Comment

<< Home