[BioRuby] Update bio-ngs, bio-gem and GSoC

[Raoul Bonnal]

Dear All,
this is an update on our activities biongs, biogem and GSoC
As you already know, BioRuby has one project accepted:
bio-objects http://bioruby.open-bio.org/wiki/Google_Summer_of_Code#Represent_bio-objects_and_related_information_with_images, Michal is the student assigned to this project.
The other candidate student Ales wants to work on his proposal bioruby wrapper http://bioruby.open-bio.org/wiki/Google_Summer_of_Code#BioRuby_Wrapper_for_Command_line_application
To optimize the time I think that we can fix a weekly meeting before or after the thursday BioRuby IRC meeting writing directly into bioruby's channel, obviously the students can write on ml and contact the mentors by irc,skype,ml etc... I think that this approach is useful also for other bioruby devs to be up to date and take part in the GSoC, because an idea/opinion is better than nothing, so your contribute is appreciated.

bio-gem: 
* I added the possibility to create an embedded database when you create a gem, some code has been borrowed from Rails and adapted to our needs
https://github.com/helios/bioruby-gem/tree/db_tasks
I'll merge this branch (db_tasks) in master, next week
Docs improved.

* rails_engines is working but not yet in master


bio-ngs & related subproject (bwa, samtools)
* wrapper is more robust now
* samtools now can be used directly from the binding https://github.com/helios/bioruby-samtools or from the wrapper. Why ? Because not all the functionalities have been bound, some time it's too complicated (see merge) So I wrote the wrapper https://github.com/helios/bioruby-ngs/blob/master/lib/bio/appl/ngs/samtools.rb
** samtools now is not shipped with precompiled library anymore, now it downloads and compiles the l library for the hist OS during the gem install ... I think that when possible we'll follow this approach.
** ricardo and dan agreed with us to have a common repository, working in that direction...
* Francesco did a great job with bwa and implementing Homology and Ontology classes and tasks to work with homology searches (i.e. blast) and Gene Ontology datasets for annotations and functional analysis. All these classes work 
with a dedicated database to store and manipulate the data.


Our abstract has been accepted for a BOSC talk, we are very happy because BioRuby will the @BOSC another time, it's a great chance to meet you and share thoughts with guys from other bio* projects.
You can download the abstract here: http://dl.dropbox.com/u/16636340/bioruby-ngs_BOSC-2011-FS-TKTYM-13-04-11-Final.pdf
They asked to talk about biogem and parallel processing as well :-) As above great chance to share ideas and experience.
We'll be @ISMB as well with 2 posters (actually we'll have a poster @BOSC too)


In these days I'm working on implementing a bio-gex class to handle gene expression datasets, it uses statsample (Claudio) to handle matrixes etc.., mostly coming from rnaseq ngs(priority) and from rtpcr, https://github.com/helios/bioruby-gex. It's an experiment but I think it could be useful, it's not a R replacement. The repo: https://github.com/helios/bioruby-gex

TODO
* write a lot of test on biongs
* refactor
* generalize some taks because we are developing them from our day by day work so it inevitably that we (mostly me, because Francesco is more precise than me :-) ) write biased code
* ....

Just a reminder for every one,  if you are going to create a new gem for bioruby please try to follow bioruby's original namespace this will help core developers to integrate plugin's code into the main repository if needed.

Cheers.

--
The only change to succeed is starting from a simple thing.