Diary of the 2003 Summer School on Theoretical and Computational Biophysics

Wednesday, June 11 -- Walter Ash

Walter Ash The day started rather typically - a faint glimmer of consciousness struggling to peel away layers of groggy sleep, motivated only by some vague sense that somewhere there might exist a ceramic object filled with bitter stimulant alkaloids in hot water...

Several minutes later, I (at this point a functional human being, thanks to the previously alluded to cup of coffee) arrived at the Beckman Institute building and took a seat in the shiny modern lecture theater. Today's agenda included a talk on lipid simulations by M. Klein (U. Penn), and an introduction to parallel scientific computing by L.V. Kale (UIUC).
"bitter stimulant alkaloids in hot water..."

The first presentation by M. Klein provided an introduction to biomembranes and lipids, including their relevance and methods used to study them at an atomistic level. A brief overview of biomembrane structure was presented, followed by a discussion of experiments used to characterize membranes and a comparison to theoretical treatments, which usually include only a very finite fraction of the complex systems (such as liposomes or oriented bilayers) studied by techniques such as NMR spectroscopy. A few case studies were presented, along with a description of some of the serious challenges still faced by scientists who model these systems.

The second lecture by M. Klein discussed coarse grained approaches to modeling biological membranes. These models, though lacking atomic detail, provide some qualitative insight into the behavior of phospholipids and other molecules interacting with membranes on time scales beyond those that are practical for atomic detail simulations. Several interesting studies on the self-assembly of phospholipid complexes (bilayers, monolayers, vesicles, etc), the behavior of water channel mimetics, and amphipathic bacteriocidal peptides were overviewed.

At this point, we headed out for our daily ration of coffee and doughnuts and (with the Beckman Institute's private garden as a pleasant backdrop) some reflection and discussion of the lecture material.

After coffee break, L.V. Kale gave a broad overview of parallel computing, followed by some case studies. Some of the challenges involved in setting up useful parallel computing platforms and designing efficient algorithms were discussed, with particular attention to problems that may arise, and some methods to overcome them. After a pleasant lunch break (tuna salad sandwich on rye) we performed some scalability tests on multiple processors using a mechanosensitive ion channel system (MscL) on the NCSA supercomputer. Finally we reconvened in the lecture hall for an overview of building inexpensive 'in-house' Linux clusters, and an opportunity to sign up for a 'hands-on' cluster building tutorial.

And after all of this, there was still plenty of time to walk to downtown Champaign for a coffee and a visit to the used bookstores and record shops.

< Previous | Index | Next >