Charla: Thermodynamic graph rewriting


Compartir
Charlista: 
Ricardo Honorato PhD Candidate University of Edinburgh
Fecha: 
4 Septiembre, 2013 - 12:00
Sala: 
Auditorio DCC
Organización: 
Éric Tanter
Bio: 

Ricardo estudió Bioquímica en la Universidad de Chile, obteniendo el grado de licenciado en 2009. Durante el ultimo año del pregrado y lo siguientes 3 años Ricardo trabajó como asistente de investigación en el Computational Biology Lab (dlab.cl). Luego comenzó un doctorado en la Universidad de Edimburgo bajo la tutoría de Vincent Danos y Gordon Plotkin en el año 2011. Hoy está terminando su segundo año de doctorado.

 
Abstract:
 
This talk introduces Kappa, a graph rewriting language to model combinatorially complex biochemical networks. In spite of being a graph rewriting language, Kappa enjoys a polynomial-time pattern matching for single connected components that enables the efficient simulation of models. Moreover, several analyses have been developed for Kappa models in the past, for instance, to estimate the set of reachable states. In the present work, we show how to statically refine a set of qualitative rewriting rules to obtain a thermodynamically consistent transition system, ie. how to obtain a finite set of rules that define an isomorphic transition system that is guaranteed to converge to equilibrium when equipped with suitable rates. The equilibrium probability is then fixed by a set of "energy graphs" with associated energy costs, which together determine the energy landscape of our system. It is important to note that the construction is entirely parametric in the energy costs (ie. the set of refined rules do not depend on them, only their rates do) and modular in the set of generator rules.