[Biopython] Phylogenetic trees with biopython?

[Paulo Nuin]

So, ClustalW trees won't even scratch the surface. You should go the  
Phylip/EMBOSS route, I don't see another alternative while using  
BioPython. Or you will have to create your own wrappers for some  
command line applications, like MrBayes, TreePuzzle, etc.

Paulo



On 1-Sep-09, at 11:17 PM, Italo Maia wrote:

> I want to use them to study evolution. Maybe guide some protein  
> modelling. Would you suggest a better use?
>
> 2009/9/1 Paulo Nuin <nuin at genedrift.org>
> ClustalW trees are extremely "simple" and only can be generated with  
> Neighbour Joining. Also they are not based on the final sequence  
> alignment created by the program but as a guide for the alignment  
> itself. They have a huge probability of being "wrong" or not  
> representing the actual relationships. It will heavily depend on the  
> the type, distance and differences among the sequences you are using.
>
> Do you need trees for what?
>
> Paulo
>
>
>
>
> On 1-Sep-09, at 10:37 PM, Italo Maia wrote:
>
> Thank you everyone for your answers. I was about to give up and run  
> my own
> wrappers over phylip(emboss for mey ubuntu is too old : /) when i  
> just found
> out that clustal can create phylogenetic trees too. On commandline,  
> with the
> options 4,1 and 4, i just made a tree here, from a .aln file  
> generated with
> clustalw. Does anyone dislike this approach? Seems like easy/fast/ 
> efficient
> enough for me.
>
> ps: i just made a simple frontend for blast, formatdb, clustalw and  
> muscle.
> Right now i'm going to add phylogenetic trees, then, i'm finished.  
> It's my
> graduation thesis, by the way.
>
> 2009/9/1 David Winter <winda002 at student.otago.ac.nz>
>
>
> David - I would prefer we also put your new wrappers in
> Bio.Emboss.Applications, and would be happy to look at adding
> those to CVS now that Biopython 1.51 is out (I had forgotten
> about them actually - so thanks for the reminder).
>
> Peter
>
>
> Hi Peter,
>
> I'd almost forgotten about them myself! I only put them in their own  
> module
> because I had the PhyML wrapper as well and that's not an EMBOSS
> application.
>
> I suspect a wrapper for PhyML is probably not going to be widely  
> useful (a
> normal run lasts at least several hours and most people will want to  
> look
> over their alignments by eye before they set it off). So I'll move the
> phylip ones into Emboss.Applications and gather a few thoughts about  
> other
> phylogenetic software including PhyML and see what the dev list  
> thinks about
> them.
>
> david
>
>
> _______________________________________________
> Biopython mailing list  -  Biopython at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biopython
>
>
>
>
> -- 
> "A arrogância é a arma dos fracos."
>
> ===========================
> Italo Moreira Campelo Maia
> Ciência da Computação - UECE
> Desenvolvedor WEB e Desktop
> Programador Java, Python
> Ubuntu User For Life!
> -----------------------------------------------------
> http://www.italomaia.com/
> http://twitter.com/italomaia/
> http://eusouolobomal.blogspot.com/
> ===========================
>
> _______________________________________________
> Biopython mailing list  -  Biopython at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biopython
>
>
>
>
> -- 
> "A arrogância é a arma dos fracos."
>
> ===========================
> Italo Moreira Campelo Maia
> Ciência da Computação - UECE
> Desenvolvedor WEB e Desktop
> Programador Java, Python
> Ubuntu User For Life!
> -----------------------------------------------------
> http://www.italomaia.com/
> http://twitter.com/italomaia/
> http://eusouolobomal.blogspot.com/
> ===========================