MetaWalks: an application to visualize metabolic pathways at different levels

Abstract

Metabolic pathways are quite difficult to visualize as a whole and to visually compare them across organisms, due to their data dimension and fragmentation. In this thesis, we propose a tool called MetaWalks to visually represent metabolic pathways at different levels, using three graph-based representations, namely, compounds and reactions graph, reaction graph and m-DAG. We employ Neo4j as a graph database and join the three representations into a single multigraph model. This allows for a great flexibility in data storing, querying and visualization, as data are stored into a single model. Neo4j is equipped with an efficient library for managing graph operations, such as the computation of the strongly connected components of an input graph. MetaWalks allows for visually comparing the topological structure of metabolic pathways of different species using the three different representations, and to easily switch among them to better explore the similarities and differences of the compared species.