Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

SOE: Fachverband Physik sozio-ökonomischer Systeme

SOE 20: Networks: From Topology to Dynamics IV (with BP, DY)

SOE 20.1: Vortrag

Donnerstag, 17. März 2011, 15:00–15:15, GÖR 226

Predicting nodes' influence in Boolean networks — •Fakhteh Ghanbarnejad and Konstantin Klemm — Bioinformatics Group, Institute for Computer Science, Leipzig University, Germany

Boolean networks serve as discrete models of regulatory dynamics in biological cells. In recent years, the Boolean modelling approach has been successfully applied to real systems. A key question in this context is the relevance of a single node for system behaviour. We quantify the dynamical influence of a node as the statistical dependence between the node's initial state and the attractor reached asymptotically. We find that the eigenvector centrality of the network is a good predictor of dynamical influence. It outperforms several other centrality measures, including out-degree, betweenness centrality and k-shell index. Quality of prediction is further improved when eigenvector centrality is based on the weighted matrix of activities rather than pure adjacencies.