D Type Documentation

Peter Gwozdz

15 Mar 2010

Summary of D type, in simple language, is available at:

http://www.gwozdz.org/PolishClades.html#DType

This "Documentation" sheet is more details

This may be difficult to understand without reviewing my methods;  follow the links in that PolishClades.html for explanations

Sheet "Haplotypes"

Row 15 is the modal haplotype definition for D type;  66 markers;  cutoff 14, gap 2

Rows 18 and 19 are 3-marker signatures for Da and Db subtypes

These subtypes cannot be resolved with statistical significance, but it is tempting to speculate there are 2 subtypes, because of the correlation of these 3 markers

Row 17 has the 67 marker modal for D type, with the three best signature markers highlighted

Sheet "Calculator"

This sheet is set up for D definition when the file opens

Result in column CR;  copy in column CL

Column BU shows the 5 samples that have been tested for DYS462 with result 12

SBP = 18.4% for any number of markers from 53 (column CI) to 61 (CH)

Because of ties in the automatic ranking, marker count jumps from 50 to 53 to 61

Best SBP = 18.1% using 66 markers

Column BY demostrates that D type is Polish

This is the data from a 17 Dec 2009 download of the Polish Project

To save file space, I deleted the samples with high step from D

But I double checked;  the definition & SBP are valid using the full Polish Project

Sheet "Ysearch" shows Ysearch resutls

Sheet "ASD" is age estimate - TMRCA

My web files & publications discuss TMRCA caveats;  these age estimates are highly uncertain

1,284 years, row 12, using all markers;  TMRCA may well be 800 to 3,000 years due to those caveats

781 years, row 17, Thomas method;  I consider this a low estimate because for type D these markers have little population structure

954 years, row 29;  this is another low estimate for TMRCA

I used a mask in row 21 to remove 8 of the 67 markers

The 464 quartet are unreliable in age estimates

The CD pair are unreliable due to obvious recLOH

578 gives a false age of 34,180 years due to only one mutation (578 is 2nd slowest of the 67), so I removed it

557 gives a false age of 9,248 years due to a single 12 value, modal 15

It is possible the TMRCA is older, because I removed too many high ones (but none of the zeros)

39 markers have no mutations, so zero age in the ASD sheet

This is additional evidence that D type is very young

The trio (458,576,444) are the signatures for hypothetical subtypes

But removing these does not produce a young cluster, because these are each rapid mutators