A Computational Framework for Learning and Memory: Network Motif Evolution During LTP-Induced Plasticity

Unraveling the complexity of network-level synaptic plasticity remains a challenge due to the dynamic and interconnected nature of neural circuits. In this study, we employ network motifs-recurrent, functionally specialized patterns of connectivity-as a framework to dissect long-term potentiation (LTP)-induced reorganization in hippocampal CA1-CA3 networks. Using network LTP recordings from highdensity microelectrode arrays (HD-MEA), we systematically tracked motif evolution before and after high-frequency stimulation, assessing their roles in network stability, synaptic strength modulation, and criticality-redundancy trade-offs, linking these dynamics to firing synchrony and network potentiation. The Graph-theoretic analysis further demonstrated that LTP-induced reorganization follows a structured motif-guided trajectory, with early-phase motif recruitment optimizing efficiency, followed by phase-dependent refinement. These findings provide new insights into how structured connectivity enables network-level plasticity, balancing efficiency with stability, and offer a potential framework for understanding memory encoding mechanisms and their dysfunction in neurological disorders.