Observer-based identification of a Multi-Output Feedforward Loop from gene expression data

In this paper, we propose an observed-based algorithm to estimate the time course of a set of not-directly measurable gene expressions for the network motif of the multi-output feed-forward loop (MO-FFL), widespread in gene transcription networks of many organisms. The MO-FFL has been modeled according to a standard ordinary differential equations approach, providing a nonlinear model in the state space. Simulations show the effectiveness of the proposed approach in a very wide range of possible critical frameworks, such as only one target gene measurements or non-smooth input perturbations.