Fig. 1: From spike trains to functional connectivity for mouse visual cortex.

A Schematic of data collection with Neuropixels probes inserted through six visual cortical areas (AM, PM, AL, RL, LM and V1), adapted from Jia, X. et al. High-density extracellular probes reveal dendritic backpropagation and facilitate neuron classification. Journal of neurophysiology 121, 1831-1847 (2019)¹²⁴. B Example spike trains of 741 units from the visual cortex of a mouse during six different types of stimuli. For brevity, each stimulus type is denoted using a unique symbol in all figures. Image credit: Allen Institute for Brain Science. [https://observatory.brain-map.org/visualcoding/]. C Example jitter-corrected CCGs (cross-correlograms) of positive (red)/negative (blue), unidirectional/bidirectional and monosynaptic (“sharp peak”)/polysynaptic (“sharp intervals”) connections. D (left) Example matrix of jitter-corrected CCG with units ordered by area during natural movie stimuli. (right) Connectivity matrix with only significant connections (∣Z∣ > 4). E Density of positive and negative connections during all visual stimuli. Density is defined as the number of connections normalized by total possible number of connections. Error bars represent 95% confidence interval, n = 7 mice. F Fraction of within-area connections, fraction of positive connections and clustering coefficient against network density. Each visual stimulus is characterized by a symbol, consistent with (B). P values are obtained through two-sided Wald test. G Kernel density estimation (KDE) of signal correlation distributions for disconnected neuron pairs and pairs with negative/positive connections during presentations of four types of visual stimuli.