Abstract Antoni Mir? Roig
In this thesis we present a systematic methodology to characterize dynamic biological systems from time series data. From the work we derived three publications. In the first we developed a deterministic global optimization method based on the outer approximation for parameter estimation in dynamic biological systems. Our method is based on reformulating the set of ordinary differential equations into an equivalent set of algebraic equations through the use of orthogonal collocation methods, giving rise to a nonconvex nonlinear programming (NLP) problem. This nonconvex NLP is decomposed into two hierarchical levels: a master mixed-integer linear programming problem (MILP) that provides a rigorous lower bound on the optimal solution, and a reduced-space slave NLP that yields an upper bound. The algorithm iterates between these two levels until a termination criterion is satisfied. In the second and third publications we developed a method that is able to identify the regulatory structure and its corresponding kinetic parameters from time series data. In the second publication we defined a mixed integer dynamic optimization problem (MIDO) which minimize the Akaike information criterion. In the third publication, we adopted a multi-criteria MIDO which minimize complexity and fit simultaneously using the epsilon constraint method in which one objective is treated as the objective function while the rest are converted to auxiliary constraints. In both publications MIDO problems were reformulated to mixed integer nonlinear programming (MINLP) through the use of orthogonal collocation on finite elements where binary variables are used to model the existence of regulatory interactions.
Download: