Groppi, F. | bioRxiv Science

bioRxiv2026

The same plasticity rule rescues networks built from one conductance set and destabilises networks built from another

Whether synaptic plasticity stabilises or destabilises a recurrent circuit is usually treated as a question about the rule. We find that in a conductance-based model it is not answered by the rule alone. The model is a simulated bursting culture: 48 excitatory and 12 inhibitory stomatogastric-ganglion neurons, sparsely and recurrently connected. The recurrent synapses carry pair-based spike-timing-dependent plasticity under a fixed homeostatic budget on each cell's total incoming excitatory conductance. We ran one net-depressing rule, at one inhibition level, on unselected random wirings, with a survival criterion fixed before the runs. Two excitatory populations built from different published conductance sets gave opposite outcomes. In one, the network collapsed without the rule in 33 of 36 wirings and the rule rescued 31 of those 33. In the other, the network survived without the rule in 11 of 12 wirings and the rule eliminated 6 of those 11. The only model parameter changed between the two is the excitatory conductance set, and that change also alters their intrinsic dynamics. We then asked what carries the reversal. Across seven conductance sets, elimination increases with a cell's own firing rate, and cell identity does not predict it: a follower-type cell firing at pacemaker rates behaves like a pacemaker. But moving the firing rate inside a single set by injected current reverses the same quantity, which rules out firing rate as a sufficient explanation. Substituting one conductance at a time from the rescued set into the other, over the six channels that differ, no single substitution transferred the rescue (2 of 42 pooled) while substituting all eight did so in 6 of 6 (Fisher exact p = 2.3 x 10^-6). The sign reversal cannot be attributed to any single tested conductance, and firing rate alone cannot account for it. All results are computational and concern one circuit model, one plasticity rule and one homeostatic constraint. Predictions and falsifiers were committed to a version-controlled record before the corresponding runs, with three departures from that order recorded in the manuscript's pre-registration index.

neuroscience↗

bioRxiv2026

Weak pre-before-post pair biases sort synaptic weights in a conductance-based model of spontaneous network bursting

Spontaneous population bursts repeatedly expose synapses to correlated spike activity, but it is unclear whether such events reinforce pre-existing synaptic weight differences when precise firing order carries little information. We asked this in a conductance-based model of a bursting neuronal culture: 48 excitatory and 12 inhibitory stomatogastric-ganglion (STG) model neurons with sparse recurrent connectivity, short-term depression, and pair-based spike-timing-dependent plasticity (STDP), with every prediction and falsifier committed to a version-controlled record before the corresponding run. Across two cohorts of 10 connectivity wirings - realizations of one model, not independent preparations - stronger synapses consistently received greater net potentiation than weaker ones, although the preregistered magnitude criterion was not met. The spike statistic most closely associated with this sorting was not first-spike latency: stronger synapses instead showed a very small but highly consistent excess of pre-before-post spike pairs, 0.2-0.35 percentage points, accumulated over approximately 10^4 pairs per synapse in 60 s (median 8,960 in the out-of-sample cohort; approximately 3 x 10^6 per wiring). In a preregistered intervention, shortening bursts by 41% at approximately matched burst rate increased the sorting measure by about an order of magnitude in all five wirings. A second, physiologically distinct perturbation reproduced the relationship with realized burst duration, but duration could not be separated from per-burst spike count. All interventional evidence is computational. We then preregistered an observational test of the corresponding prediction in 16 archived recordings of developing cortical cultures. The predicted strength-ordered pair excess was not detected, and the predicted negative relationship with burst duration was absent. The former is a bounded null - the recordings exclude effects above +0.32 percentage points but cannot resolve the model's approximately 0.2-point magnitude; the latter was adequately powered under the preregistered criterion. A consistent positive latency-strength relationship, unpredicted by the model, was observed. The study therefore establishes a reproducible sorting phenomenon within a specific computational model and a testable burst-shortening prediction, but does not establish that the mechanism operates in living cortical networks; an intervention on cultured networks with an independently validated measure of synaptic strength is the decisive test.

neuroscience↗