Accepted_test

Computational biology widely uses mathematical models of biological systems, in particular, systems of ordinary differential equations, as well as models of dynamical systems described in other formalisms, including discrete models such as finite automata, Petri nets, Boolean networks, and individual-based models. Predicting changes in the type of solution dynamics as a function of changes in model parameters is a substantial scientific problem. However, this problem does not have analytical solution for most formalisms in a general case. Widely used practice of conducting a series of computational experiments, i.e. solving a series of direct problems with different sets of parameters followed by expert analysis of solution plots, becomes arduous with a large number of parameters and decreasing step of the parametric grid. Thus, developing analytical methods that are capable of working on a set of computational experiments in an aggregated form is of eminent importance. A method of visualisation and classification of different dynamic regimes of the model was developed using a composition of dynamic time warping algorithm and principal coordinate analysis. Time series ordination for computational experiments dynamics (TSOCED) visualises the results of multiple solutions of a mathematical model in the space of reduced dimensionality. TSOCED also allows to reveal a correspondence between the parameter values and the type of dynamic modes of the model’s solutions.