Things I noticed #8

All the recent changes have made my life in the blogosphere a little erratic of late. That said, here is the latest set of things I noticed this past week

Cross-correlating at Foo Camp

The O’Reilly Radar has a post on presentations at the recent Foo Camp that define personify convergence. One was a presentation on visualizing change logs for Wikipedia and the other on visualizing SNPs. Like Tim O’Reilly, I found myself quite fascinated by the striking similarity of the visual representations. while, like one of the commenters noted, this is somewhat of an artifact, it doesn’t make it any less fascinating. It also made me aware of a rather brilliant gentleman by the name of Ben Fry

The Connectivity Map

I am not sure I like the name all that much, but Bioinform (subs required) alerted me to the publication of a connectivity map by researchers at the Broad Institute. The map is a proof of concept and currently consists of “564 gene expression profiles derived from screening 164 small molecules under several different cell lines”. The idea behind the map is somewhat obvious. In theory users can look at various signatures of drug response at the transcript level. As the map gets expanded, it will be interesting to see how access is controlled and how it gets used. I also think that over time it will not be limited just to gene expression. The paper describing the map was published in Science. This subject is interesting enough to merit a detailed follow up and I hope to do so shortly.

A new model for publication and content?

Also from Bioinform, is the news of the announcement of the Targeted Proteins Database (TPdb), a database that adds layers of information to entries on various proteins in SwissProt. The database comes from Current BioData, and will be followed by a family of open-access journals related to protein families covered by the database. It will be interesting to see if the protein information itself is opened up. Perhaps there should be a curated version and a “raw” version, one similar to wikipedia, with oodles of information, automated, added by dedicated “miners” and by users. The idea of journals connected to a database is an interesting one in itself and the involvement of BMC definitely suggests that the plan (the database is due for release in 2007) is a serious one. I will be following this story as it unfolds.

Protein folding on graphics cards

Keeping with the whole “reported in Bioinform” theme, Vijay Pande and co-workers at Stanford have taken their Folding@Home project to a new frontier, using graphics cards for the distributed computing effort (this after PS3 support was announced a few months ago). Currently supported on ATI graphics cards, this is a result of ATI’s recently announced Stream Computing technology, which is a natural prgression for GPUs, which have excellent floating point capabilities. Obviously, there are many more challenges (as the Bioinform article covers quite well), but its a good place to start.

HP joins the healthcare club

Our friends at Bioinform also report that like IBM, Intel, and Microsoft, Hewlett-Packard is setting up a formal group dedicated to healthcare and life sciences. Given how much they play in this space, it is a little surprising that one didn’t exist before, but they have enough connections and presence in the sector that this step only formalizes the effort.

The EU has a semantic grid project

Genetic Engineering News (to get away from bioinform for a bit) had a press release on the EU adopting Inforsense for their semantic grid project. The interesting bit for me was the fact that the EU had such a project in place. The project, called ARGUGRID is part of the 6th Framework program (FP6) and is a general purpose framework for modeling the grid at a “semantic knowledge-based level of abstraction”. I wonder if there are any similar, centrally funded, efforts in the US?

The future of online storage

Everyone’s favorite Yahoo blogger, Jeremy Zawodny, has a very interesting analysis of how he has started using Amazon’s S3 service as a personal backup system. He goes through a lot of the details of the cost benefits and I must say, it is a sign of the times. Right now, bandwidth and network speeds are the major bottleneck, but when the average person has 100 mbps and higher upload times and reliable connectivity, I don’t see why the majority of people won’t use online services for backup storage. The costs will certainly come down as the use increases.

Postgenomic meet Google Gadgets

When Google opened up Google Gadgets, the favorite science blog resource in these parts had one of its own.

Technorati Tags: Data Content, Databases, Data Storage, Publication, Protein Folding, GPUs, Graphics Cards, Distributed Computing, Folding@Home, Visualization, Foo Camp, EU, Semantic Wed, Grid Computing, HP, Healthcare, Life Science, Computing, High Performance Computing, Scientific Computing, S3, Amazon, Postgenomic, Google Gadgets

powered by performancing firefox