GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-12 > 1197169515


From: "Alfred A. Aburto Jr." <>
Subject: Re: [DNA] figuring modal values
Date: Sat, 08 Dec 2007 19:05:15 -0800
References: <e2f96c1b0712071913x5b7d1f90r4223a226712dc444@mail.gmail.com><e2f96c1b0712081125i1a8f71abpe79f4e8dacf9a415@mail.gmail.com>
In-Reply-To: <e2f96c1b0712081125i1a8f71abpe79f4e8dacf9a415@mail.gmail.com>


Robert,
Something to note about the "taken singularly" cases is that the
histograms have structure. Both CDYa & CDYb show multiple peaks
indicating subclades or sub-modal characteristics. I would say that
CDYa & CDYb are actually bi-modal. CDYa shows two peaks at (33,36) and
CDYb shows two peaks too at (36,38). The CDY combinations you get from
this are CDY=33-36, CDY=33-38, CDY=36-36, and CDY=36-38. Two of these
are the same as the "taken together" and the other two are different. I
think though with more data that the modal characteristics of the "taken
together" and "taken singularly" will line up better ... At least this
this is something you can check on as you gather more data ...
Al


> Robert Tarín wrote:

>Thanks to all who have replied. I have also had some discussion off-List
>about this and it was pointed out that in the Chandler study "Estimating
>Per-Locus Mutation Rates" http://www.jogg.info/22/Chandler.pdf on the 37
>marker FTDNA set only DYS389 is separated. Other multi-copy markers such as
>DYS385, DYS464, CDY, YCAii are taken as a whole. I suppose it may be due to
>the difference in rates between DYS389-1 (0.00186) and DYS389-2 (0.00242),
>but that leaves the question are different markers treated differently based
>on mutation rates of their individual repeats? It seems RecLOH events would
>mess things up.
>
>I also understand more data is better but wonder if there could be
>situations that mirror instances such as this one which is based
>upon real-world data. Sample size is 36 in this example.
>
>CDYa,b (taken together)
>31-36 n=2 (5.56%)
>32-39 n=1 (2.78%)
>33-33 n=2 (5.56%)
>33-34 n=2 (5.56%)
>33-35 n=1 (2.78%)
>33-36 n=3 (8.33%) modal
>33-38 n=1 (2.78%)
>34-34 n=2 (5.56%)
>34-36 n=2 (5.56%)
>34-37 n=1 (2.78%)
>34-38 n=1 (2.78%)
>35-36 n=3 (8.33%) modal
>35-38 n=1 (2.78%)
>35-40 n=1 (2.78%)
>36-36 n=2 (5.56%)
>36-37 n=3 (8.33%) modal
>36-38 n=3 (8.33%) modal
>36-39 n=1 (2.78%)
>36-43 n=1 (2.78%)
>37-38 n=1 (2.78%)
>38-38 n=2 (5.56%)
>
>CDYa (taken singularly)
>31 n=2 (5.56%)
>32 n=1 (2.78%)
>33 n=9 (25.0%)
>34 n=6 (16.67%)
>35 n=5 (13.89%)
>36 n=10 (27.78%) modal
>37 n=1 (2.78%)
>38 n=2 (5.56%)
>
>CDYb (taken singularly)
>33 n=2 (5.56%)
>34 n=4 (11.11%)
>35 n=1 (2.78%)
>36 n=12 (33.33%) modal
>37 n=4 (11.11%)
>38 n=9 (25.00%)
>39 n=2 (5.56%)
>40 n=1 (2.78%)
>43 n=1 (2.78%)
>
>In the above scenario you would have different results for modal values.
>Taken together as a set you end up with a quad-modal 33-36, 35-36, 36-37,
>36-38. Individually you get 36-36 as modal which is not one of the
>quad-modals.
>
>Robert
>
>


This thread: