 Hannah
|
North_Dakota_Wheat
|
Jacob
|
Friday, May 07, 2004
9th Structural Biology Symposium
This last weekend I attended the Structural Biology Symposium in Galveston Texas. It was sponsored by the Sealy Center for Structural Biology and hosted by the University of Texas Medical Branch (UTMB). Dr, Adam Arkins presented some very interesting work that merged the use of game theory and genomic network modeling to simulate the advantages of strategies like chaotic (looked like a domain around a strange attractor from classical chaos theory) mutation requirements that enhance the ability of an organism to adapt to changing environmental pressures. That is the behavior of this organism's biochemical oscillators (as modeled by negative feedback) when taken as a consistent single characteristic would likely fail (become extinct) given a change in environmental stress. How ever when the organism's gene expression has a chaotic tendency, then a subset of the population would have the more robust character ( like possible going to the sporulation stage sooner) required for the species to bridge over that external stress.
Wolfgang Baumeister presentation on the state of art in Electron Tomography was fantactic. With the current resolution of 8 nm many larger protein complex are visible in vivo. The correlation that he demonstrated was between 75% and 88% matching given the signal to noise ratio of the experiment. A high resolution model from other approaches could be "de-res'd" to provide a signature that the program could search and match. I don't quite remember the three proteins that he was using to bench mark the performance, but I think one was GroEL and the other two were also barrel structured proteins of similar size.
This last weekend I attended the Structural Biology Symposium in Galveston Texas. It was sponsored by the Sealy Center for Structural Biology and hosted by the University of Texas Medical Branch (UTMB). Dr, Adam Arkins presented some very interesting work that merged the use of game theory and genomic network modeling to simulate the advantages of strategies like chaotic (looked like a domain around a strange attractor from classical chaos theory) mutation requirements that enhance the ability of an organism to adapt to changing environmental pressures. That is the behavior of this organism's biochemical oscillators (as modeled by negative feedback) when taken as a consistent single characteristic would likely fail (become extinct) given a change in environmental stress. How ever when the organism's gene expression has a chaotic tendency, then a subset of the population would have the more robust character ( like possible going to the sporulation stage sooner) required for the species to bridge over that external stress.
Wolfgang Baumeister presentation on the state of art in Electron Tomography was fantactic. With the current resolution of 8 nm many larger protein complex are visible in vivo. The correlation that he demonstrated was between 75% and 88% matching given the signal to noise ratio of the experiment. A high resolution model from other approaches could be "de-res'd" to provide a signature that the program could search and match. I don't quite remember the three proteins that he was using to bench mark the performance, but I think one was GroEL and the other two were also barrel structured proteins of similar size.
Comments:
Post a Comment
