Collins Conference Room
Seminar
  US Mountain Time

Our campus is closed to the public for this event.

Eric Deeds (University of Kansas)

Abstract. Protein interaction networks are critical for life. Cells use these networks to process information in order to respond adaptively to stimuli in their environment. Protein interactions are also central to the assembly of a host of “macromolecular machines” that perform key cellular functions. Our work focuses on using mathematical and computational models to understand the dynamical properties of these networks. We recently constructed a model of the pheromone signaling cascade in yeast; this network allows yeast cells to sense and mate with nearby cells. It is widely (if often implicitly) believed that systems like the pheromone cascade function via the formation of specific “signaling machines,” representing well-ordered complexes that propagate the signal. The pheromone network is, however, a highly combinatorially complex system: it can theoretically generate over a billion distinct protein complexes. Using rule- and agent-based simulation techniques explicitly designed to handle combinatorially complex systems, we found that this network can function without the need to form a specific signaling machine. Instead, information can be transduced through the formation of a highly heterogeneous ensemble of complexes. This finding indicates that cellular computations can occur through purely local interactions in the absence of significant long-range order. There are, however, a number of “true molecular machines,” like the ribosome and the proteasome, which do indeed seem to function via the formation of a single, highly ordered macromolecular structure. We have developed efficient simulation techniques to model the assembly of these types of complexes. We have found that such machines can readily suffer from what we term “assembly deadlock,” a type of kinetic trapping in which large intermediates persist after all the monomers and other small building blocks are exhausted from the system. Recent experimental work from our collaborators has confirmed that deadlock can indeed occur in the self-assembly of macromolecular machines. Together, these findings represent a starting point for understanding the general principles of complex formation, and thus function, in the protein interaction networks that underlie cellular life.

Purpose: 
Research Collaboration
SFI Host: 
Jennifer Dunne

More SFI Events