Dice Question Streamline Icon: https://streamlinehq.com

Foundational open questions on neural connectomes

Determine whether cellular-scale neural connectomes are complex networks exhibiting scale-free degree sequences and the small‑world property; Ascertain the extent to which neuron spatial positions are determined by network topology (i.e., whether spatial patterns can be inferred from the identity of network neighbors); Evaluate whether, given fixed neuron positions, the wiring of the connectome is near‑optimal with respect to spatial cost; and Develop generative network models that jointly capture the principal topological and spatial features of neural connectomes.

Information Square Streamline Icon: https://streamlinehq.com

Background

The paper analyzes cellular-scale connectomes from fly, mouse, and human volumetric reconstructions and introduces associated contactomes (physical contact networks) to quantify spatial constraints. The authors frame their paper around unresolved foundational issues at the level of neural connectomes, concerning complexity (degree distributions and small‑world features), the interplay between topology and spatial embedding, wiring optimality, and the feasibility of generative models that unify topological and spatial constraints.

These questions motivate the development and evaluation of maximum entropy models that incorporate degree sequences, spatial distance dependencies, and physical contact constraints, with the goal of understanding shared organizational principles across species.

References

Yet, there are several key open questions to address at the level of the neural connectome, such as: i) Are the neural connectomes complex networks in the usual sense, showcasing a scale-free degree sequence and the small-world property? ii) To what extent is the position of a neuron given by its neighbors, i.e., are the spatial patterns determined by the topology? iii) Are there signs of optimal wiring given the neuron positions, i.e., to what extent is the connectome dictated by the spatial aspects? iv) Can we design generative network models that capture the main topological and spatial features of neural connectomes?

— Combined topological and spatial constraints are required to capture the structure of neural connectomes (2405.06110 - Salova et al., 9 May 2024) in Introduction