Synaptic Plasticity Shapes Triplet Connectivity Motifs in Spiking Networks

Authors: Claudia Cusseddu, Dylan Festa, Christoph Miehl, Julijana Gjorgjieva

Presentation type: Talk at SNUFA 2025 online workshop (5-6 Nov 2025)

Abstract

Neural circuits process and interpret information through a continuous interplay between connectivity and activity. In spiking neural networks, this relationship is shaped by precise spike timing, with spike-timing-dependent plasticity (STDP) refining the network connectivity. Of particular interest are connectivity motifs—small subnetworks of a few neurons—that link structural features of the circuit to dynamic properties such as covariability and dimensionality. Experimental studies, from multi-patch recordings to large-scale connectomic reconstructions, have found nonrandom distributions of three-neuron (triplet) motifs across brain regions and species. However, the plasticity mechanisms underlying the formation of motifs remain poorly understood. In this work, we investigate how triplet connectivity motifs emerge in spiking networks using a parametrized family of STDP rules. By considering a range of biologically plausible rules — including Hebbian, anti-Hebbian, and symmetric — we explicitly model how spike-timing interactions drive synaptic changes. We develop an analytical framework that links the parameters of such spike-based plasticity rules to triplet connectivity motifs, allowing us to predict which motifs can emerge from any given STDP rule. Our theoretical predictions are validated by large-scale simulations of spiking networks. Our results show that no single STDP rule can account for all observed triplet motifs. However, combining distinct STDP rules facilitates the formation of all possible triplet motifs. Importantly, we demonstrate that simple combinations of plasticity rules are sufficient to reproduce the heterogeneous motif distributions observed experimentally. In conclusion, our work reveals how synaptic plasticity shapes triplet connectivity structures, highlights the importance of heterogeneity in models of biological circuits, and provides a foundation for future investigation into the interplay between plasticity and dynamics.