Our research efforts are focused on understanding actin dynamics and force generation using a combination of analytic theory and computer simulation. Actin polymerizes rapidly in response to external signals, and often displays an autonomous dynamic behavior. Several of the key molecular-level processes underlying actin dynamics are known, and these include polymerization/depolymerization of filaments, branching/debranching, capping/uncapping of filament ends, and bundling of filaments. Our current focus is understanding how actin dynamics in endocytosis is determined, and how actin polymerization provides pulling forces to drive the process. In our simulations, we retain the coordinates of all actin subunits in filaments over time, thus obtaining a detailed picture of the three-dimensional structures that form, such as branched networks and bundles. Recently, we have shown that  pulses of actin polymerization that occur during endocytosis in yeast result from a negative-feedback interaction in which actin polymerization pulls actin nucleators off the membrane. Furthermore, we found that pulling forces in endocytosis are generated by a characteristic profile of actin polymerization where the polymerization is slowest at the edges. We are currently trying to understand how actin filaments can generate the large forces required to overcome the turgor pressure in yeast.