[Biojava-l] some comments and wish list

[Bruce Ling]

Hi,

As for the future of biojava, I have some wish list here.

I think bioperl has been extremely successful as it has been really
tailored for the majority of the bioinformaticians at this time.  The task
facing the field really focuses on genomic sequence and the annotation
manipulation.  Unfortunately biojava lags behind and doesn't really provide
a lot of tools for that aspect.  If I am going to use biojava to process
blast, HMMER or some other databases files, basically I have to spend a lot
of time to understand the  biojava code and reinvent a lot of the wheels. I
have to say there is not many building block which can be fetched out of
the shelf of biojava.  Documentation and demos code are nice but not
enough.  -Some of the demos code are nice, but it is very crude and really
cannot be used for large scale data processing.  Is it possible that
biojava can spend more efforts on that aspect.  I think the most critical
issue here is how to provide easy to use API such that details of the
implementation can be hidden.  I like bioperl because they did a terrific
job on that aspect.  I guess to transform the functionality of the bioperl
(because of its good design and functionality) is a must if you want to
have more traffic toward the biojava. I see one good direction of the
biojava is its xml parser development, but there is no development in the
code which constitutes bio seq objects directly from the parser or from seq
objects to xml.  One question, are you going to support GAME in biojava in
the near future?

Any way, just some random thoughts.  Hopefully it can help the leaders in
biojava to plan for the future projects.

Bruce Ling