Data for Brain Reference Architecture of NY24CanonicalCorticalMicrocircuitsInference

Canonical cortical microcircuits reference architecture for cognitive inference

抄録

Canonical cortical microcircuits (CCMs) is the six-layer structure preserved throughout the mammalian neocortex and is thought as a fundamental computational unit. This dataset reverse-engineers CCMs and presents a computational model to achieve cognitive inference. The data consist of the anatomical connectivity of CCMs and the functions hierarchically achieved from each uniform circuit. First, information on the anatomical connections of CCMs was collected from seven review papers. Next, dynamic Bayesian inference was determined as the algorithm for cognitive inference, which CCMs can implement. Finally, we describe how top-level functions are achieved from excitatory neural populations and circuit motif based on inhibitory neural populations, assigning output semantics to each excitatory neural population. The data are described in a brain reference architecture format and stored in the BRA data repository. This dataset provides experimentally testable hypotheses about neural activity patterns in cortical layers.