Tuesday, July 31, 20122:00 PM - 3:30 PMCNLS Conference Room (TA-3, Bldg 1690)|
q-Bio Seminar Series
New methods for modeling stochastic biochemical networks
Eric MjolsnessEric Mjolsness UC Irvine Departments of Computer Science and Mathematics
I will discuss the theoretical derivation of several new computational methods for modeling stochastic biochemical networks developed in my group. These include "ER-Leap", an exact acceleration of Gillespie's stochastic simulation algorithm; a hierarchical version of ERLeap which can be parallelized; the "Graph-Constrained Correlation Dynamics" method for model reduction, which seeks to maintain a Boltzmann distribution approximation to the probabilistic system state; and a method for learning the parameters in stochastic chemical kinetic models.
Joint work with David Orendorff, Todd Johnson, Yuanfeng Wang, Scott
Christley, and Xiaohui Xie.
Host: S. Gnanakaran, 5-1923, firstname.lastname@example.org