COMPUTATIONAL MODELLING OF CONNECTIVITY ESTABLISHMENT IN THE AUDITORY PATHWAY

Across sensory systems, the development of topographic maps provides a canonical framework for understanding how connectivity is established and refined into precise wiring. While this phenomenon has been classically studied in retinotopic projections, the axonal tract linking the cochlear nucleus (CN) to the medial nucleus of the trapezoid body (MNTB) pathway offers a complementary system presenting tonotopic mapping and near one-to-one connectivity. How can topographic and topological precision emerge from initially overabundant CN-to-MNTB contacts? As in other sensory relays, initially exuberant CN-MNTB contacts are refined by patterned spontaneous activity and competition. We present a CN-MNTB simulation framework combining probabilistic formation of candidate contacts, patterned spontaneous activity, and activity-dependent synaptic competition with pruning. Candidate contacts are generated with a probability that depends on geometry and tonotopic mismatch, and synaptic weights are initialized from this probability. We then simulate structured spontaneous activity in the CN along the tonotopic axis and update synapses using a Hebbian-like potentiation term together with a competitive, homeostatic constraint on total input to each MNTB neuron, followed by removal of persistently weak contacts. Preliminary simulations indicate that, across plausible parameter settings, spontaneous activity driven refinement improves topographic order and reduces excess connectivity. Overall, this work provides an interpretable framework to generate testable hypotheses about how probabilistic arborization and spontaneous activity with competition shape pruning dynamics and the emergence of tonotopic and topological precision in auditory brainstem circuits.