Analysis of Biochemical Pathway Perturbations with Network Propagation

In this paper we propose a network propagation algorithm tailored to Chemical Reaction Networks. The approach is designed to tackle the problem of understanding how alterations in the kinetic constants of reactions propagate inside a biochemical pathway affecting each species taking part in the biological process. The desiderata is to work exploiting just the topology of the network, without having access to the parameters governing the dynamics required by other solutions that model the evolution of species concentrations by means of deterministic and stochastic algorithms. The solution has been tested on a dataset of synthetic pathways in order to evaluate its performance and spot problematic features of the pathways that lead to worse accuracy scores. Finally, it has been also tested and evaluated on a set of real pathways to see whether the results obtained in the artificial cases generalize to real ones.