Neuroscience is a fascinating subject for many, and for various reasons, I have a rather soft spot for the field. The computational challenges of neuroscience are, surprise surprise, rather intriguing. After all, our brain is in many ways the ultimate computer, and we are always trying to figure out how it works and how to replicate intelligence artificially.
Through Frank Gibson’s screencasts on Bioscreencast, I found out about the CARMEN, a neuroinformatics project that aims to “create an e-science infrastructure in which data on neuronal activity (electrical and optical measures) can be shared, stored, manipulated and modelled”.
Neuroinformatics is data intensive and is fed by heteregenous data types. Unfortunately a lot of the data is confined to a single lab, i.e. the broader community is not being tapped properly (something that was forced onto biology by the sheer breadth of the human genome project). But the CARMEN project leapfrogs some of the trajectory of bioinformatics, by providing a virtual laboratory that incorporates knowledge about experimental conditions and extensive amounts of meta-data. It took biofinformaticians a long time to realize that capturing experimental information as meta-data was important.
CARMEN is an ambitious project, and to be truly successful will need to last a lot longer than the current 4 year pilot.
This presentation tells you about the CARMEN project, and some of the challenges the project hopes to address.
Recent Comments