後頭頂葉皮質は、同じメタ構造だが逆の行動系列を共有する経路に沿った進捗をマッピングする

Path networks defined by patterns of intersections and orientations among individual paths may contain recurrent meta-structural features that can be learned and used in spatial reasoning and navigation.¹ Posterior parietal cortex (PPC) neurons exhibit scalable tuning to progress through routes,² suggesting that PPC in rats, as in humans, may function to encode common structure among multiple routes.^3,4,5,6 PPC neurons were recorded as rats performed a spatial working memory task within a network organized by recurrence in the layout of short and long pathways and right angle intersections. The specific routes utilized provided opportunity to contrast responses of PPC sub-populations to navigational actions with more complex responses that map route progress and features of route structure. We found evidence for the presence of PPC neurons that generalize in their firing patterns across routes having the same shape but opposite action series. A separate population discriminated the same routes and exhibited activity more closely related to angular velocity. The results indicate that PPC has the capacity to generalize the mapping of route progress and meta-structural organization in parallel with encoding of navigational actions. We suggest that such encoding can form the basis for learning the meta-structural organization of a non-random path network structure, such as that commonly found in cities. Further, the findings suggest that PPC could compute relationships between higher-order spatial information and navigational action sequences.