5.1 Introduction

This final chapter of the thesis is devoted to improve the network representation and analysis of metabolic networks.

The two main topics around which the chapter is structured are:

• How to accurately mathematically describe metabolic networks as graphs and how to include the maximum amount of available biological information in those graphs.

• How to use those graphs for a simple E. coli metabolic model (the same used in Chapter 4) and how use community detection methods to find an insightful biological network structure of the bacteria.

The first two sections (Summary and Background) address the main objectives of the work carried out in this chapter and the basics of graph representations of metabolic networks.

The third section (Methodology) explains briefly the building blocks necessary to achieve the proposed goals.

Those are mainly Flux Balance Analysis (see Section 4.5.2 for a detailed description) and Markov Stability community detection method (see last subsection in Section 2.3.1).

However, most of the construction of the new graphs is placed in the Results and discussion section.

The fourth section (Results and discussion) shows the bulk of the work done.

First, the construction of the matrices and graphs is thoroughly described.

Then, the graphs for the core E. coli model are built for several different biological scenarios or conditions.

Finally, the community structure of these networks is analysed.

The last section (Conclusions) draws the main conclusions from this chapter.