The interplay of synchronization and fluctuations reveals connectivity levels in networks of nonlinear oscillators

We study spatiotemporal patterns produced by small-world networks of biologically motivated nonlinear oscillators from a data-analysis perspective. It is shown that the connectivity levels of such systems can be reconstructed by analyzing heterogeneity and fluctuation content of the patterns. These properties are determined by applying spatiotemporal filters described in [Physica A 289 (2001) 498] to pairs of oscillators in a network. Possible applications of our method to biological data (e.g., time-resolved cDNA microarray data), in order to distinguish densely connected systems from sparsely connected systems, are commented on.