Signs, polynomials, and reaction networks

Introduction

Many real-world problems are reduced to the study of polynomial equations in the non-negative orthant, and this is in particular the case for models of the abundance of the species in a biochemical reaction network. The polynomials associated with realistic models are huge, with many parameters and variables, making qualitative analyses, without fixing parameter values, challenging.

This results in a mismatch between the needs in biology and the available mathematical tools. The driving aim of this proposal is to narrow the gap by developing novel mathematical theory within applied algebra to ultimately advance in the systematic analysis of biochemical models.

Research Focus

Motivated by specific applications in the field of reaction networks, we consider parametrized systems of polynomial equations, and address questions regarding:

1. The number of positive solutions.
2. Connected components of semi-algebraic sets.
3. Signs of vectors.

Specifically, we:

1. Pursue a generalization of the Descartes' rule of signs to hypersurfaces, to bound the number of negative and positive connected components of the complement of a hypersurface in the positive orthant, in terms of the signs of the coefficients of the hypersurface.
2. Follow a new strategy to prove the Global Attractor Conjecture.
3. Develop new results to count the number of positive solutions or find parametrizations.

Novelty and Strength

The novelty and strength of this proposal reside in the interplay between the advance in the analysis of reaction networks and the development of theory in real algebraic geometry.

The research problems are studied in full generality for arbitrary parametrized polynomial systems, but each question has a well-defined purpose, directed to the ultimate goal of having a scanning tool to automatically analyze the models used in systems and synthetic biology.

Therefore, this proposal will strengthen the bridge between applied algebra and real-world applications, through the study of reaction networks.