Notes from the Biomass

The International Max Planck Research School for Computational Biology and Scientific Computing is a joint graduate program of the Free University of Berlin and the Max Planck Institute for Molecular Genetics, Berlin.
Its 3-year PhD program starts in autumn 2006 and is open to students from inside and outside Germany. The deadline for applications is February, 28th.

Experimental Study of Inequality and Unpredictability in an Artificial Cultural Market[s] is the title of a new study by Salganik, Dodds and Watts published in Science today. It is the perfect piece of publishing in our artificial cultural market containing everything to maximize public impact: pop culture, social networks and a bit of high-school math, showing why experts routinely fail to predict anything.

It would be the perfect forward pass to start a long rant on pop star scientists, scale free networks and the Arctic Monkeys if it wasn't a thought through study delivering unexpected results.

Damn.

[Pop-cultural coverage at Telepolis[de] and Seed magazine]

Old Nodalpoint has always been more than a multiuser blog and recently polished its forums. Incidentally, my recent search for useful discussion groups on bioinformatics subjects turned out to be futile again. Now I hope that Nodalpoints user base will ensure the quality of the posts. There is really not much point to discuss the same subject in every science blogs in solipsism.

Back in 2002, I would not have believed that Wikipedia could work out. You all know that it did became a major success but recently concerns were voiced, including quality issues, sometimes due to intentional addition of misleading information or simple deletion of unwanted facts, and the lack of experts on particular subjects.

Now, the Digital Universe wants to create a resource for peer reviewed scientific information, backed by Larry Sanger[w], who worked on Nupedia, a predecessor of Wikipedia and funded by Joseph Firmage, a victor of the New Economy who maintains an odd proximity to the UFO community. Digital Universe wants to secure the quality by stewardship of experts to particular research fields and boasts to become the "largest reliable information resource".

The first item on their roadmap already smells quite odd.

1 - Browser Independence
First, we’ve built the first version of the Digital Universe to work with the Mozilla-based ManyOne browser. In a few months, you’ll be able to access the Digital Universe from any popular browser, and also use text-based navigation in addition to visual navigation, if you prefer.

Several months to achieve browser independence? In 2006?

Nature[s] covers the story in its current issue but you might also want to check the The Register[f], which voices concerns on both projects - and the reply and corrections from Larry Sanger.

Wikipedia is currently aiming to attract scientists as lack of experts, particular in the natural sciences, is being discussed in the forums extensively. Sanger is definitely right in that Wikipedia is not a very appealing place for experts right now but there are major efforts to change that. However, the mix of experts and lay people should result in articles that focus on readability and understanding for the majority of users rather than expert opinions, an interaction that I have enjoyed. I have my doubts whether Digital Universe can compete with the quality initiative in Wikipedia and finally work out - but let's wait for the first articles.

The recently launched Biology Direct differs from the vast majority of scientific journals in that it publishes under an open peer review scheme. The names of the reviewers and their reports are published alongside the original publication. Announced in October 2005 it focusses on genomics and bioinformatics and is backed by an impressive line up of scientists. The first publications went online yesterday, accompanied by a editorial that underlines the experimental nature of the journal.

Launching a new research journal in biology in the year 2006 takes a lot
of hubris…and/or a clearly defined goal. The crucial open access niche has been taken by
the highly successful and still proliferating BMC and PLoS journals, so a new journal
hardly would stand a chance and be worth the efforts of the editors and the publisher
unless it defines itself in a fundamentally new way. Thus, our goals with this new journal,
Biology Direct, are unapologetically ambitious: to establish a new, perhaps, better system
of peer review and, in the process, bolster productive scientific debate, and provide
scientists with useful guides to the literature.


Are peer reviewed publications really a good way to foster scientific debate? Peer review is beneficial to the point of being unmissable for primary research publications but I am not convinced that it enhances debates. For Biology Direct, the facility to comment on the work is included for most BMC titles anyway, though I fail to see widespread use of it, and I don't see where Biology Direct provides additional means. A system that would not require peer review but imposes more structures than blogs, wikis of news forums could provide more to discussion by many different scientists - not that I would know what exactly to implement.

I had hoped for publications with higher profile - the four current papers are somewhat interesting but I would have expected that Landweber, Lipman, and Koonin would have been able to pull higher impact papers into the new journal. Probably, most scientists submitted rather solid but not controversial work and reserve their high profile papers for journals with higher impact (or with an impact factor to begin with).

Still, the direction that Biology Direct will take in the next couple of years will tell us how flexible the publishing scheme really is. PLoS changed quite a bit of it already and there's undoubtedly more coming.


[via F&L]

EMBO journal carries a serious survey [f] of how communication between scientists erodes over time due to "decay" of email addresses. Their survey is based on Medline, analyzed over the last ten years.

One in four e-mail addresses becoming invalid within one year of publication is an alarming rate of decay as it has an impact on the ability of scientists to communicate and exchange material.

The analysis includes - amongst typos and organizational problems - the factor that many email addresses are abandoned voluntarily - unhindered communication might be seen as a nuisance also. Recommended (and it's free).

Delivering random presentations from the internet is an idea I've heard and discussed a couple of times as past time activity at scientific workshops. We never actually went as far as selecting the slide shows and presenting them though. Now, after visiting a public show recently, I realized that Powerpoint Karaoke is both good fun and useful training and recommended activity for retreats or a night with the department's beamer in the seminar room.

You need to prepare a selection of slide shows (filetype:ppt in Google and a random keyword works for me) but that should be good fun already. Agree on or impose a voting system, the less objective, the better, just to make sure that people don't take it overly seriously.