Notes from the Biomass

Many German bioinformaticians usually view the yearly German Conference on Bioinformatics as a necessary evil. Who wants to listen to mediocre talks in stale run down German university lecture theatres? - the sole reason to go was networking. However, I (and many others) consider RECOMB and ISMB - the two major bioinformatics conferences - to have a similarly disappointing line up compared to smaller workshops with a high degree of invited speakers.
The way these conferences are set up currently is just not very rewarding and while there are many submissions, the number of promising submission is fairly low. If bioinformaticians have to watch their impact factors (unlike say mathematicians), submitting to a conference is not worth the hassle, even if some of the talks are published in Bioinformatics.
Luckily, this years talks - contributed and invited - were much better than in previous years and I am glad I attended. The organization was smooth (owing to the increased registration fee, some people say) and the main building of the university of Hamburg was a pleasant venue and Hamburg, is a metropolitan city worth a visit anyway. Tübingen, certainly nice but not exactly a metropolis, will host the next GCB from September 22 to 24th, 2006.

Many people criticized the current state of publishing in the life sciences, in particular Eugene Koonin, whose views on peer review and the long delays that go with it made it into Nature. The launch of Biology Direct, a new BioMed Central journal for everything around biology and bioinformatics (with an initial focus on the latter) offering href="http://www.biology-direct.com/default/10078-aims.htm">novel open peer review might actually change the system. This is going to be very interesting - the launch of PLoS forced many publishers to go open access - let's see what happens to the ill-mannered closed peer review.

The German Bioinformatics conference that I am currently attending has a regional focus, China. Guo-Ping Zhao from the Chinese National Genome Center, Shanghai started the day with an overview of the genomics involved in the outbreak of SARS in 2003. It was amazing to see how quickly comprehensive information such as the genomic sequence on the agent was obtained and how genotyping of the SARS-strains could be matched back to most individual cases of infections. The structural biology of several SARS proteins was investigated by Rolf Hilgenfeld, who was working on Coronaviruses before. Parts of the work that he presented - the solution of the crystal structure of the main protease of the virus - was actually performed in China in a university under quarantine.
Jin Li, director of the newly founded partner institute between the Chinese Academy of Sciences and the Max-Planck-Society presented an overview of recent developments in the HapMap-project, focussing on the insights into population genetics, Using haplotype blocks, they were able to reconstruct the major migration of the human population. It will be interesting to listen to people who criticized the project early on when The Big Paper appears, which Jin Li announced.
Obviously, the people presenting ideas on normalization of microarray data had a tougher time attracting attention. However, as long as data normalization plays such a pivotal role, we have to go a long way before microarray will deliver their potential, despite their abundant use today.

Another fine day here - both in the lecture theatre and outside.

The first day of the GCB had two sessions, "RNA structures" and "Sequence Analysis". Peter Stadler himself called the first session "dominated by the RNA mafia", referring to the group around Peter Schuster that Stadler and the other speaker were part of at some point in time.
Their work is probably known to everyone in RNA folding and evolution by the Vienna RNA package . Nice work and talks but I have no real use for studying RNA folding myself. RNA folding is one of good examples where staying in a field that was pronounced dead (when everybody was doing genome sequencing) proved to be very fruitful (when small RNAs appeared).

The highlight of the second session was the biodefense presentation by Tom Slezak from the Lawrence Livermore National Laboratory. The problem posed is interesting: how to find the best DNA regions of pathogens that distinguishes pathogenic strains from their avirulent cousins. The results are monitored by PCR for rapid diagnosis but also for surveillance on air filters. I would have hoped for a little more information but some of "the customers" - federal agencies - are secretive about many details. Outside the biodefense application, it would be nice to get information on the "air metagenome".

More tomorrow.

Today, the GCB, Germanys most important bioinformatics conference opens in Hamburg. The venue, the university of Hamburg, has a strange touch to it - this was the location of my very first "conference". Not on saving the world from bugs using computers but on fantasy role playing games - back in 1989. Well, I won't get to sleep on some floor this round and with WLAN set up, expect some posts on the conference later.

The Nobel Prize in Medicine 2005 is awarded to the Australian scientists J. Robin Warren and Barry J. Marshall for the discovery of Helicobacter pylori.

Marshall infected himself with the bug to prove Koch's postulate, creating a lasting impression on me (and giving me the cramps really). It was also one of the first bacteria that we fully sequenced because there was little known about it 10 years ago. The Helicobacter field received copious amounts of grant money after the initial discovery, making it one of the research fields of bacteriology in the late 90s.

Genomic sequencing of prokaryotes is now in a stage where it's no longer sufficient just to sequence the genome of a prokaryote with some relevance for publication in a good journal. You will have to come up with novel insights and either an interesting bug (getting rarer...) or sequence a bunch of related genomes. The latter was performed by a group at TIGR on Group B Streptococci, reporting that sequencing a few strains are not sufficient to represent all strains in the species.

From the abstract:
Analysis of these genomes and those available in databases showed that the S. agalactiae species can be described by a pan-genome consisting of a core genome shared by all isolates, accounting for approximately 80% of any single genome, plus a dispensable genome consisting of partially shared and strain-specific genes. Mathematical extrapolation of the data suggests that the gene reservoir available for inclusion in the S. agalactiae pan-genome is vast and that unique genes will continue to be identified even after sequencing hundreds of genomes.


Dr Ferfried Gutfind from the Barbara Cartland Center for Positivistic Genomics says: "The massive sequencing of six strains is a beautifully analyzed, yielding the surprising conclusion that species diversity in some niches will not be covered ever by traditional sequencing approaches. Even "classical" meta-genomics won't help to cover the sequence space in these organisms."
However, Noland Works, chairman of "Bioinformatics students against basically everything" concludes: "Yeah, roight, after sequencing six strains theses guys say that they need to sequence way more species to ever reach their goal in a contributed paper to PNAS. Sounds like setting up publications for references if you ask me." (I won't.)

See also the note in The Scientist.

"Math and Science Song Information, Viewable Everywhere", a repository for science songs, from the whacky to the ... err ... not so whacky.

I do miss Air's Sex Borne Poison, featuring the Prince from the Biomass.

[Via Riesenmaschine (in German)]

Back in Germany and with a couple of days passed, the meeting can be called on of the best ones I attended so far. In short, the quality of the talks (and many posters) was high, the organization smooth and the location very inspiring anyway.
The abstracts books lists about 250 participants from all over the world, not surprisingly, US based scientists were in the majority. Talks were usually well attended with no parallel sessions. A strict time limit was enforced for the individual contributions (10 or 30 minutes) but there was no time set for the next speaker, allowing flexibility for questions. All nights, we concluded after 10pm, thereafter one could mingle with the participants at the one bar. The venue enhances the interaction - all activities focussed in a single lecture theater, the dining hall and the bar, all very close by each other, and speaking to anybody was easy this way.

And the science?
This was a meetings for experimental microbiologists with relatively little immunology (phew) and no bioinformatics, which was to be expected from the speakers list. It would not be fair to report details from a closed meeting, even if the CSH press policies does not list blogging in particular. Greg uttered some concerns on closed meetings; I do think that such meetings foster co-operations and mutual trust rather than providing only some cheap thrill of exclusivity. Several speakers provided insights into current work that I felt glad to see early and presented openly - not all work was in the process of submission.
The next CSH meeting on Microbial Pathogenesis and Host Response will be from September 15th to September 18th, 2007. However, there are many other meetings held at Cold Spring Harbor in other fields that might be more interesting to bioinformaticians. Venue and organization can definitely be recommended.