[Biojava-l] Interested in working with BioJava for GSoC 2012

Hannes Brandstätter-Müller biojava at hannes.oib.com
Sun Mar 18 11:10:15 UTC 2012


While the ideal solution would be a clean port of the BLAST algorithm
to Java, the complexity of such a mature and performance-optimized
code might be a bit too much for a GSOC project.

The JNI solution, while not optimal, would be an acceptable fallback
solution. If implemented properly, a base for using existing programs
like BLAST, HMMER, … via JNI could be created.

The project goal is: "Make BLAST work locally, without internet
connection in addition to the existing NCBIQBlastService method"

Hannes

On Sun, Mar 18, 2012 at 08:42, Komal Sanjeev <komalsnehal1991 at gmail.com> wrote:
> Hi Amr,
>
> The following has been mentioned in the project description:
> "Converting C or C++ source code by hand is not a trivial undertaking and
> it is recommended that a C/C++ to Java conversion tool be used to do as
> much of the work as possible. It is also an option to consider a JNI
> approach for integrating these applications into Java."
>
> I am a bit confused. Kindly tell me what exactly has to be done in the
> project.
>
> Thanks,
> Komal
>
>
>
> On Sun, Mar 18, 2012 at 9:35 AM, Amr AL-Hossary
> <amr_alhossary at hotmail.com>wrote:
>
>>   Dear Komal,
>>
>> As far as I know, The project is about porting the code to Java, not using
>> existing C code within JNI.
>> That means you should be able to digest the algorithm first, Build it
>> using Java from scratch, depending on C code as a reference Implementation.
>>
>> Regards
>>
>> Amr
>>
>> -----Original Message-----
>> From: Komal Sanjeev
>> Sent: Sunday, March 18, 2012 12:59 AM
>> To: Andreas Prlic ; biojava-l at lists.open-bio.org
>> Subject: Re: [Biojava-l] Interested in working with BioJava for GSoC 2012
>>
>> Hi all,
>>
>> Introducing myself a bit more, I also work as a remote intern for
>> DARNED<http://darned.ucc.ie/>.
>> DARNED is a database of RNA Editing, and currently new features are being
>> added into the project, one of which is incorporating the BLAST feature for
>> sequence based search. Having worked on a similar project recently, I think
>> I will be comfortable working with the 'Porting an Algorithm to Java'
>> project.
>>
>> The  following is what I understood about the project. Please correct me if
>> I am wrong.
>> This(link) <http://biojava.org/wiki/BioJava:CookBook3:NCBIQBlastService>
>> is
>> the current method used for BLAST, which accesses the NCBI website each
>> time. The NCBIQBlastService class is currently used. The project aims at
>> replacing this with code which will perform the search within Biojava. I
>> downloaded the source codes of BLAST and HMMER. My job will be to convert
>> these to Java.
>>
>> Regarding the C/C++ to Java converter, i found this on the internet:
>>
>> http://tangiblesoftwaresolutions.com/Product_Details/CPlusPlus_to_Java_Converter_Details.html
>>
>> but it is not free of cost.
>> Apart from this, I saw that many people discourage the use of C/C++ to Java
>> tools saying that they are not efficient. Does anyone know of any better
>> tool which can do this?
>>
>> Regarding the JNI, would it not be better if the whole code was written in
>> Java, rather than a part of it being in C/C++? I haven't used it before,
>> but if it is better than converting the code, I don't have a problem
>> working with it.
>>
>> Kindly clear my doubts.
>> Thanks in advance,
>> Komal
>>
>>
>>
>> On Thu, Mar 15, 2012 at 1:17 AM, Andreas Prlic <andreas at sdsc.edu> wrote:
>>
>> > Hi Komal,
>> >
>> > stay tuned to this list, we still don't know if we will get funded
>> > from Google this year.
>> >
>> > Andreas
>> >
>> >
>> > On Wed, Mar 14, 2012 at 12:27 PM, Komal Sanjeev
>> > <komalsnehal1991 at gmail.com> wrote:
>> > > Hi Everyone,
>> > >
>> > > I am Komal, an undergraduate student from IT-BHU, India. I'm interested
>> > in
>> > > working with BioJava for GSoC 2012. I am particularly interested in
>> > working
>> > > on the 'Porting an Algorithm to Java' project.
>> > > Kindly help me about how I should proceed.
>> > >
>> > > Thanks,
>> > > Komal
>> > > _______________________________________________
>> > > Biojava-l mailing list  -  Biojava-l at lists.open-bio.org
>> > > http://lists.open-bio.org/mailman/listinfo/biojava-l
>> >
>> _______________________________________________
>> Biojava-l mailing list  -  Biojava-l at lists.open-bio.org
>> http://lists.open-bio.org/mailman/listinfo/biojava-l
>>
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biojava-l




More information about the Biojava-l mailing list