Dominant patterns in small directed bipartite networks: ubiquitous generalized tripod gait

Abstract

The synchronization patterns exhibited by small networks of neurons that regulate biological processes (CPGs) have aroused growing scientific interest. In many of these networks there is a main behavioral pattern within the parameter space. In particular, in the context of insect locomotion, tripod walking stands out as a predominant pattern, both in natural observations (where insects walk on tripod gait) and in mathematical models. This predominance appears to be stable under parameter variations within the network, suggesting a possible correlation with the underlying network topology. Tripod walking can be naturally extended to all CPGs with a bipartite connectivity. Then a natural question arises: Are “generalized tripod gaits” equally dominant among synchronization patterns within those networks? To investigate this, we carried out a comprehensive study covering all bipartite networks of up to nine neurons. For each of those networks we numerically explore the phase space using a quasi-MonteCarlo method to see what are the main synchronization patterns that the network can achieve. Then, all those patterns are grouped according to their dynamics. Generalized tripod gait was observed in all cases examined as the dominant pattern again. However, certain cases revealed additional stable patterns, mainly associated with the 3-colorings of the respective graph structures.