Modularity and robustness of bone networks

Matheus Palhares Viana,Esther Tanck,Marcelo Emílio Beletti,Luciano da Fontoura Costa

DOI: https://doi.org/10.1039/b814188f

2009-01-01

Molecular BioSystems

Abstract:Cortical bones, essential for mechanical support and structure in many animals, involve a large number of canals organized in intricate fashion. By using state-of-the art image analysis and computer graphics, the 3D reconstruction of a whole bone (phalange) of a young chicken was obtained and represented in terms of a complex network where each canal was associated to an edge and every confluence of three or more canals yielded a respective node. The representation of the bone canal structure as a complex network has allowed several methods to be applied in order to characterize and analyze the canal system organization and the robustness. First, the distribution of the node degrees (i.e. the number of canals connected to each node) confirmed previous indications that bone canal networks follow a power law, and therefore present some highly connected nodes (hubs). The bone network was also found to be partitioned into communities or modules, i.e. groups of nodes which are more intensely connected to one another than with the rest of the network. We verified that each community exhibited distinct topological properties that are possibly linked with their specific function. In order to better understand the organization of the bone network, its resilience to two types of failures (random attack and cascaded failures) was also quantified comparatively to randomized and regular counterparts. The results indicate that the modular structure improves the robustness of the bone network when compared to a regular network with the same average degree and number of nodes. The effects of disease processes (e.g., osteoporosis) and mutations in genes (e.g., BMP4) that occur at the molecular level can now be investigated at the mesoscopic level by using network based approaches.