[Taxacom] Fwd: Re: base alignment

pierre deleporte pierre.deleporte at univ-rennes1.fr
Mon Aug 28 11:12:58 CDT 2006



Hi John,

if I may I suggest two foreseable difficulties for your understanding...

- below you're again talking of "cladistic" for a question of alignment, 
i.e. a priori homology assessment (= "primary homology" assessment). Many 
discussions on this list have shed light on the fact that "cladistic" is 
generally applied to the analysis of the data (= parsimony analysis), not 
to constituting the data matrix (like alignment). And our off-list 
discussions allowed me to understand that what you mean by "cladistic 
characters" is the set (or 'clique' in jargon) of completely 'compatible' 
characters (= homoplasy-free characters) as resulting from a previous 
compatibility analysis ('clique analysis') at a higher phylogenetic level - 
even if you perform this preliminary analysis "in your brain" and not with 
the help of a computer, you can get the correct result in very simple cases 
(involving few taxa).

- as for models, even a classic cladistic analysis on morphological 
characters (not 'cladistic characters', but parsimony analysis these 
characters) implements a "model", generally implicit rather than explicit, 
and frewuently named "descent with modification" (but it is more specific 
than this: descent with modification and no common mechanism of character 
change etc...). Only some strict "pattern" cladists pretend to infer a 
phylogeny without implementing evolutionary models for character change 
(which I find logically inconsistent, just like pretending to perform 
evolutionary biogeography without any evolutionary assumption).

but my principle hint is simply that "starting from scratch" requires your 
adjusting your vocabulary to the current usage; hence I recommend as starters:
"The Compleat Cladist" (E.O. Wiley et al.), and "Basics of cladistic 
analysis" (Diana Lipscomb), both available for free on the web.

and of course most of Victor Albert's "Parsimony, Phylogenetics and 
Genomics" is highly recommendable... but only after you have fixed a 
minimal basic common language.

e.g., if you look after sdome evocatiopn of your own "cladistic 
characters", you'll have to seek after "compatibility analysis" in Albert 
et al., and you'll just get some mathematical developments for character 
analysis (of homoplasy-free characters) in Steel and Penny's paper - 
"cladists" in this book visibly do not perform compatibility analysis on a 
daily basis. Some developments of Dollo analysis could also be of some 
intererest for you, but homoplasy is allowed in DSollo parsimony 
(reversals), hence this would likely not taste enough of "cladistic 
characters" for you.

and if you think of "homoplasy-free characters" everytime you read the term 
"cladistic", you'll likely be completely lost, because "cladistic" is never 
applied to characters (except by yourself it seems, as far as I know).

if this can help... but please note that these "basics" in terminology are 
in no way specific to molecular data analysis, they apply to morphological 
data as well, hence your likely lasting difficulties for understanding 
cladistic litterature and discussing with your colleagues for morphological 
data and analyses as well.

Best,
Pierre


>A 09:20 25/08/2006 -0400, John Grehan wrote :
>>In the course of continuing my investigations on the orangutans and
>>human origins evidence I am beginning to delve more into the molecular
>>claims. This is an area I do not pretend any expertise (on the other
>>hand when I started I didn't have any expertise with primate morphology,
>>but that did not prevent me finding out that many of the claims for
>>derived human-chimpanzee characters were either unsubstantiated or
>>undocumented) so I have to start from scratch.
>>
>>I would be interested in any feedback or direction to suitable
>>publications or web sites that outline the procedures by which
>>substitutions and gaps are balanced to result in base alignments between
>>species. In my current (and perhaps primitive) understanding, this
>>process does not appear to be cladistic, but a theoretical model by
>>which homologies are constructed. If this understanding is correct (and
>>so some of you may point out that it is not) it would seem to me that
>>character states in molecular studies are not empirical in the same way
>>that they are in morphology (i.e. the homologies are the result of a
>>calculation in molecular studies).
>>
>>John
>>
>>  Dr. John R. Grehan


Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 63
Télécopie : 02 99 61 81 88


_______________________________________________
Taxacom mailing list
Taxacom at mailman.nhm.ku.edu
http://mailman.nhm.ku.edu/mailman/listinfo/taxacom




More information about the Taxacom mailing list