Producing a Set of Models for the Iron Homeostasis Network

TL;DR

This paper introduces a novel modeling approach combining formal interval techniques, simulations, and STL satisfaction to address parameter uncertainty in biological systems, demonstrated on the iron homeostasis network.

Contribution

It presents a new method for modeling biological systems that handles large parameter uncertainties through formal interval analysis and systematic parameter space exploration.

Findings

Produced a robust model of the iron homeostasis network

Explicitly described the regulation mechanism at the translational level

Validated the approach on biological data

Abstract

This paper presents a method for modeling biological systems which combines formal techniques on intervals, numerical simulations and satisfaction of Signal Temporal Logic (STL) formulas. The main modeling challenge addressed by this approach is the large uncertainty in the values of the parameters due to the experimental difficulties of getting accurate biological data. This method considers intervals for each parameter and a formal description of the expected behavior of the model. In a first step, it produces reduced intervals of possible parameter values. Then by performing a systematic search in these intervals, it defines sets of parameter values used in the next step. This procedure aims at finding a sub-space where the model robustly behaves as expected. We apply this method to the modeling of the cellular iron homeostasis network in erythroid progenitors. The produced model describes explicitly the regulation mechanism which acts at the translational level.