[BioRuby] Blast to Phylogeny

Tomoaki NISHIYAMA tomoakin at kenroku.kanazawa-u.ac.jp
Thu Nov 19 00:33:32 UTC 2009


Hi,

In general, to construct a phylogenetic tree from molecular sequence  
data,
you will collect the homologous sequences, perform multiple alignment,
identify the region that will be used for the reconstruction,
and then pass the data to an appropriate program to reconstruct the  
phylogeny.

If I have a BLAST output, I would parse that file with Bio::FlatFile and
extract the identifiers of the hit sequences, use the identifiers to  
collect
individual sequences and submit the sequences to mafft for multiple  
alignment.
Convert the alignment to nexus format and manually check with  
MacClade, and then
parse the edited nexus file to write the multiple alignment readable  
by the phylogenetic
analysis program. There are many options you can take at each step.

So, there are multiple ways, but not a single simple way. :(

Bioruby has support for multiple alignment programs like mafft,  
muscle, and clustalw.
For phylogenetic reconstruction, there is some support for phylip and  
paml
(I don't have tried these feature from Bioruby library, though).
There are a number of programs for phylogenetic analysis other than  
phylip and paml.
A list compiled by J. Felsenstein is available at
http://evolution.genetics.washington.edu/phylip/software.html

An alignment similar to that of phylip will be accepted by most  
programs.
-- 
Tomoaki NISHIYAMA

Advanced Science Research Center,
Kanazawa University,
13-1 Takara-machi,
Kanazawa, 920-0934, Japan


On 2009/11/19, at 5:16, Sharvari Gujja wrote:

> Hi,
>
> I am trying to construct a phylogenetic tree from Blast  
> output...Could you please let me know if there is a way to do  
> this..I have also been looking at Bio::Tree documentation but it is  
> not clear if it accepts Blast file as input.
>
> Appreciate any help.
>
> Thanks
> Sharvari




More information about the BioRuby mailing list