Connectomics

Connectomics is the large-scale mapping of synaptic connectivity in neural tissue — the project of determining, at nanometer resolution using electron microscopy, which neurons connect to which, through which synapses, with what structural characteristics. It is the most ambitious attempt to produce a complete wiring diagram of a nervous system.

The field's flagship achievement to date is the complete connectome of Caenorhabditis elegans: 302 neurons, 7,000 synapses, mapped by White et al. in 1986 and refined computationally since. Knowing the complete wiring diagram of C. elegans has not, in four decades, fully explained how the worm generates its behavior. This is the fundamental lesson of connectomics: the wiring diagram is a necessary but not sufficient condition for understanding a neural circuit. The same wires carrying different signals produce different behaviors; neuromodulation changes which connections are functionally active without altering the physical wiring.

More recent efforts include the partial connectome of a cubic millimeter of mouse cortex (Microns project, 2021) and human cortical columns. These are engineering achievements of the first order — producing petabytes of image data and requiring machine learning pipelines to segment individual neuronal processes. What they are producing, structurally, is a graph of extraordinary complexity with unknown dynamical properties. The map is not the territory. neural dynamics — the time-varying activity patterns that implement computation — cannot be read from a static wiring diagram.