The plan was to create separate new BJ3 jars for task-specific code, exactly as you suggest for phylogenetics. I'd support a biojava-phylo jar of some kind, and I agree it would probably depend on the BJ3 biojava-core module for sequence handling. The existing BJ code was not originally going to be refactored into separate jars, unless Andreas has other plans!
Scooter Willis wrote: > Andreas > > Congrats on taking on the responsibility of steering BioJava in a > positive direction. > > I needed the ability to generate phylogenetic trees from aligned > sequence data and found that the work was started in a google summer > project but looking at the code it wasn't finished and appeared to focus > only on loading trees not creating them. I ended up taking the tree > generation code out of jalview and removing as much jalview dependencies > as possible and have it as a nice tight collection of classes. My > assumption without any deep legal review is that because jalview is open > source that the code can be used and contributed to another open source > project like BioJava. I will also plan on contributing the code changes > back to Jalview. > > One of the challenges I ran into with the JalView code is performance > for building the tree when using 1800+ sequences(takes a very very long > time) so I am doing some code optimization and finishing up testing on a > fairly significant performance speedup doing Neighbor_Join with a > slightly different approach that makes it N2 instead of N3. I have a > couple things to fix in tree joinging code and then will compare results > for the quality of the tree compared to the original distance matrix. I > should know more this week. > > I think I remember a BioJava discussion about trying to seperate parts > and pieces to that if you try and use a particular feature set of > BioJava you are not forced into absorbing the entire BioJava collection > of Jars. In my case I would want a biojava-phylogenetic.jar that has all > things related to tree creation and/or tree viewing etc. If the common > data format for handling sequences is RichSequence or Sequence then I > would expect to have one other Jar requirement of biojava-core.jar. Not > sure if any work has been done to refactor the BioJava code base into > multiple jar files in the same way apache does its jars for great java > code geared to a specific problem domain. > > Let me know what I can do to assist moving forward. > > Thanks > > Scooter Willis > > On Tue, Apr 14, 2009 at 4:33 AM, Richard Holland > <[email protected] <mailto:[email protected]>> wrote: > > Hello again. > > Well, nobody objected, and several people supported the idea, so I would > now like to formally hand over control of the BioJava project to Andreas > Prlic with immediate effect. > > It's been good fun working with the project over the last 5 years, and > although I'll no longer be in charge, I will still remain on the mailing > lists and contribute code/ideas/bugfixes/etc. whenever I get the chance. > > I'll also continue to attend BOSC, including this year in Stockholm, so > I'm looking forward to meeting up with everyone there for a beer or two. > > Thanks for the help and support everyone's given, and I'm sure you'll > join me in wishing Andreas the best of luck with the project. He'll be > an excellent leader and with him in charge I believe the project will go > from strength to strength. > > cheers, > Richard > > -- > Richard Holland, BSc MBCS > Finance Director, Eagle Genomics Ltd > T: +44 (0)1223 654481 ext 3 | E: [email protected] > <mailto:[email protected]> > http://www.eaglegenomics.com/ > _______________________________________________ > Biojava-l mailing list - [email protected] > <mailto:[email protected]> > http://lists.open-bio.org/mailman/listinfo/biojava-l > > -- Richard Holland, BSc MBCS Finance Director, Eagle Genomics Ltd T: +44 (0)1223 654481 ext 3 | E: [email protected] http://www.eaglegenomics.com/ _______________________________________________ Biojava-l mailing list - [email protected] http://lists.open-bio.org/mailman/listinfo/biojava-l
