Computational model of developmental self-assembly of a cortical microcircuit

Calibrated BioDynaMo simulations self-assemble ~70,000 cortical neurons from 1,200 progenitors, reproducing experimentally measured connectivity across 64 connection types with mean absolute error 0.0062.

• Why it matters: Large-scale models often prescribe wiring instead of explaining how cortical connectivity forms from local developmental interactions.
• What they did: BioDynaMo models migration, neurite growth, and contact-based synapse formation in 3D physical space, using one calibrated filling fraction per connection type.
• The result: The resulting spiking networks show stable, inhibition-dominated activity, generate both a canonical and visual-cortex microcircuit under different targets, and a reeler-like column when migration is inverted.

AI summary of the abstract · may contain errors · not medical advice · read the paper.