Meeting Review: The Bioinformatics Open Source Conference 2001 (BOSC 2001)

The general aim of the BOSC was to bring people up to date with what is happening in the open source community, which has representatives all over the world. Some of the talks were only five minutes long (Lightning Talks). These were a great idea as they gave many more people the opportunity to introduce a new area or update on an old one. However, their short length did mean that a certain level of knowledge was presumed. BioPerl had perhaps the biggest presence but as the oldest Bio* (BioStar) that is not really surprising, although BioPython was close on its heels, with applications in many areas. Much of the open source collaborations are supported by the Bio* project using hardware provided by Sun and hosted by the Genetics Institute in Boston, USA. A general cry for more people to be involved with the coding seemed to be a common lament of all the Bio* projects. All in all an enjoyable if rather quick (only two half days) and intense look at what’s happening in the BOS Community. More details as well as abstracts of the talks and posters are available from http://www.open-bio.org/bosc2001/. Steve Brenner from UCA Berkley started the conference with a talk on The Open Source Author’s Contract. This was aimed at the Americans in the audience with unfortunately, no feed back from the different European countries as to what the equivalent situation is here, although it definitely gave us all food for thought. He pointed out that creating open source software at an American university is usually illegal, as the university often owns the copyright! Something not many of us had realised. Apparently, this is similar to the situation in many British universities. He also put forward that there should be a standardised contract for coders to use that superseded university policy and which should be signed as a condition of employment, with the legal fees paid for by the open Bio consortium. This would be the responsibility of the group leader who could opt for open source, commercial or both so long as any sponsors agreed. Thomas Down (Sanger Centre), introduced us to BioJava, a relative newcomer (in comparison to BioPerl) although there are already a large number of libraries available, including some for BLAST parsing (http://org.BioJava.utils.query). Andrew Dalke (Dalke Scientific Software) gave an update on BioPython: Status and Plans. BioPython looks very interesting and seems to be quite an intuitive programming language. Python is a high level language like Perl, it is runtime, readable and easy to use. There are BLAST parsing capabilities; an SQL interface to BioPerl DB, and may be used for search and retrieval from e.g. GenBank (http:// www.biopython.org/). Jason Stajich (Duke University) presented a talk on BioCORBA. CORBA (Common Object Request Broker Architecture) is the industry standard to connect (wrap) different types of tools and databases, irrespective of what language they are written in. BioCORBA is the Bio* implementation of this. Recently BioCORBA wrappers have been written for EnsEMBL and EMBOSS. A BioPerl binder has been written to make BioCORBA Objects into BioPerl or BioJava objects and vice Comparative and Functional Genomics Comp Funct Genom 2001; 2: 327–329. DOI: 10.1002 / cfg.107