RE: Using "P" in a Cladistic Engine

["Toby White" <mwhite@houston.rr.com>]

PHYLLIP is an awkward and probably obsolete engine, but it's *free* and
it does handle polymorphisms well.  I usually use the DNAPARS feature
which permits up to six monomorphic states (a,c,g,t,u,and -).  The good
part is that it also accepts separate designations for *any* combination
of 2-3 morphs a,c,g,and t:

            M       a or c
            R       a or g
            W       a or t, etc...

So, unless you have very complex polymorphisms, or absolutely must use
more than 6 character states, it works quite nicely.  That is, when
tested against PAUP, I get almost the same answers.  I suppose the small
differences may be due to the inaccurate codong of polymorphies under
PAUP.

Toby White
http://www.palaeos.com 

> -----Original Message-----
> From: owner-dinosaur@usc.edu [mailto:owner-dinosaur@usc.edu] 
> On Behalf Of Jaime A. Headden
> Sent: Friday, November 01, 2002 1:50 PM
> To: dinosaur@usc.edu
> Cc: Mickey_Mortimer111@msn.com
> Subject: Using "P" in a Cladistic Engine
> 
> 
> 
> --- Mickey Mortimer <Mickey_Mortimer111@msn.com> wrote:
> 
> <2. Frontal process of premaxilla extending posteriorly to 
> level of lacrimal Lu et al. miscoded oviraptorids (p), 
> enantiornithines (p) and Hesperornis (0) (see above).>
> 
>   I just want to jump on this in case anyone's doing it in 
> their matrices:
> 
>   Coding "P" as a "polymoprhic" condition in a character 
> state acts _as_ a character state in the matrix. There is no 
> way in PAUP to use a separate character aside from that 
> assigned to be missing "?", a gap "-", or a series of varying 
> states "1, 2," etc.. "P" is included in the last group and 
> PAUP treats it as a valid state, meaning that any taxon with 
> it will run a tree where it tries to group with other taxa 
> having the same state coded as so. I noticed that recently 
> some recent matrices of the Theropod Working Group Matrix and 
> our own Tom Holtz have resolved to using actual different 
> states in parentheses, which can include any number of 
> states, e.g. "(01)" or "(134)" and PAUP treats each occurance 
> of a state in a different tree. Some may wish to simplify 
> this method by assigning each variable another character, say 
> x for "(012)" but this, as in "P", gives a false signal of 
> assignment, and PAUP actuallys runs the tree with "(01)" as 
> if the same position was occupied by a "0", and then occupied 
> by a "1".
> 
>   Variable transformation and distribution of a condition in 
> any taxon is usually broken down when taxa are also broken 
> down to just their species or "genera", as has become the 
> preferred method in recent matrices (see for instance Xu et 
> al., 2002; Hwang et al., 2002; Holtz, 2001, with some 
> exceptions, and there are others; genetic matrices are always 
> based on selection from individuals, but these use a slightly 
> different version of matrix when in PAUP that permit the four 
> states available ["A, C, G," and "T"] only). It thus seems 
> problematic that assigning a "P" condition, which is what is 
> being done, will screw up any matrix at least a little, and 
> the more "P"'s in the matrix, the worse your signal is likely to be.
> 
>   Cheers,
> 
> =====
> Jaime A. Headden
> 
>   Little steps are often the hardest to take.  We are too 
> used to making leaps in the face of adversity, that a simple 
> skip is so hard to do.  We should all learn to walk soft, 
> walk small, see the world around us rather than zoom by it.
> 
> __________________________________________________
> Do you Yahoo!?
> HotJobs - Search new jobs daily now
> http://hotjobs.yahoo.com/
> 
>