Program
| MSC3a | |
|---|---|
| Jon Alberts | |
| University of Washington | |
| Title | Agent-based simulations of filaments, crosslinkers, and motor proteins |
| Abstract | Complicated cellular behaviors emerge from networks of simple molecular and/or force-based interactions, and they underlie all of biology. Our failure to understand such "emergent phenomena” limits progress today in many areas of research. We therefore urgently need new methods for dealing with such complexity. One very promising tool is so-called agent-based computer modeling, which involves the explicit simulation of small-scale local interactions, following the trajectory through time of thousands to millions of states. The construction and application of these models integrates classical mechanics with biophysics and computer science, and it requires a great deal of data from experimental cell biology. Detailed agent-based simulations give us a way to explore, through a blizzard of arithmetic on fast memory-laden computers, the complex emergent behaviors that characterize all interesting cellular behaviors. This mimicking of biological systems in silico does not generate an elegant mathematical encapsulation of a system. But it has the great advantage of avoiding any need to intuit the outcome of myriad biochemical and force feedback loops, a task at which human intelligence is demonstrably frail. I will demonstrate features of a 3-dimensional agent-based modeling framework, describe biophysically realistic actin filament and myosin motor representations, and present results from the modeling of networks of filaments, crosslinkers, and motors. |
| Location | Woodward 3 |