[Bioperl-l] Building consensus sequence from ESTs

[Elia Stupka]

> I don't know about GCG ... and I doubt if there anything within bioperl
> to do this, but here are my options (please enlighten me with other
> variations): 

Hello Fernan,

this is just a brief comment to say that we are working to write the
wrappers in bioperl for some of the tools you mentioned, most importantly
we want to be able to interface biopipe (www.biopipe.org) with StackPack,
so we are likely to start by having a top-level wrapper that runs the
StackPack process from start to end, and later we are going to work with
EG folks to try and wrap individual components so that some of its parts
(like the phrap stage) could be pipelined over the cluster.

Nothing productive to give out as yet, but just to keep you and other in
the loop about our direction.

Elia


> 
> We work with ESTs and we have used different approaches to EST
> clustering and the production of consensus sequences. 
> 
> i) phrap or any other sequence assembler would cluster ESTs, and also
> produce consensi. Many parameters available to play with. 
> 
> ii) STACKpack, from the South African National Bioinformatics Institute
> and Electric Genetics. Free for academic use. It is a full package, with
> a MySQL db, web && command-line interfase, etc. It uses several other
> tools (cross_match/RepeatMasker, phrap), but comes with one that makes
> the difference from other clustering approaches, it uses as a first
> clustering step a loose comparison tool d2_cluster to form initial
> clusters that are then assembled using phrap. Produces consensus and has
> a tool to analyze them and suggest alternative forms (splicing?)
> http://www.sanbi.ac.za/CODES/
> 
> iii) compare ESTs all vs all using BLASTN and group sequences using these
> similarity relationships. Then align them using clustalw and get your
> consensi. This approach would benefit from bioperl (running BLASTs and
> parsing results, generating the groups, running clustalw on them and so
> on).
> 
> iv) perhaps EMBOSS has an app similar to the GCG's Gelmerge you
> mentioned? If so, you can use bioperl interfase to EMBOSS to automate
> the process.
> http://www.emboss.org/
> 
> Regards, 
> 
> Fernan
> 
> 

Elia

********************************
* http://www.fugu-sg.org/~elia *
* tel:    +65 6874 1467        *
* mobile: +65 9030 7613        *
* fax:    +65 6779 1117        *
********************************