Archiver > GENEALOGY-DNA > 2010-02 > 1265456498

From: "Irish III DNA" <>
Subject: Re: [DNA] Some haplotypes that had a problem on Whit's predictor
Date: Sat, 6 Feb 2010 22:41:38 +1100
In-Reply-To: <6BBE62C0045D4CD1B02F5BF0FDA67229@PC>

Andrew wrote:-
I have these from a recent survey. I can say more generally that Whit's
predictor had big problems with most J1a divisions with this marker set, but
perhaps the J1a splitting in Whit's predictor is a bit ambitious for this
number of markers.

The sequence is

These ones seemed surprising though, because the decision is between
phylogenetically distant clades...

1. The following two haplotypes were predicted as either E1b1b or J1, with a
slight favour to J1:-

Being more familiar with E1b1b, I am surprised J1 can be so E like?

But I am wondering if the H4 might be a conversion issue.

2. The following was either R1b or T:-

There definitely seems to be a problem with the H4 values. Conversion normally is only ? 1 but the range is only 7-14 so 20 or 21
doesn't fit.
Have you tried Jim Cullen's predictor?

It doesn't have a slot for C4/635 but without this marker and H4, it produced the following:-

The first two sequences it shows as:-
E3b =>42% E3b-M34 =>27% J1-M267 =>14% E3b-V12 =>9% E3b-V22 =>5% I-P37.2 =>1%
and E3b =>44% E3b-M34 =>21% E3b-V12 =>18% E3b-V22 =>7% J1-M267 =>6% I-P37.2 =>1%

The third is:- R1b-Irish/Continental =>47% R1b-S.Irish =>26% R1b-Frisian4 =>14% I-P37.2 =>4% R2-M124-Cluster-B =>3%
R1b-C.Europe =>2% Ix-S23 =>1% R1b-Ub =>1%

Dennis Wright

This thread: