Crosstalk measures for analyzing biological networks in breast cancer

Understanding the interaction and crosstalk between pathways is important for understanding the function of both physiological and pathological biological systems. We have taken a computational approach to explore interactions among modules within biological networks by comparing and contrasting various topological measures which may be useful in the identification and prediction of critical connectivity points between modules. Node degree, betweenness, bridges, and articulation points may define connections among modules with distinct functions. Structural holes are another topological feature of networks which are important in identifying the role of nodes in the relationships among subclusters of graphs. Structural holes separate non-redundantsources of information, sources that are more additive than overlapping. We explore the performance of these among protein-protein interactions in yeast, then apply them to gene networks derived from a cohort of early stage breast cancer patients in whom different levels of IGF ligand have been associated with differing outcomes. We compare the different approaches to identifying and ranking genes based on these measures to reveal clues about cross-talk and feedback mechanisms and their role in mediating communication and coordination among modules.