Finding the neural basis of consciousness is challenging, and the distribution location of the core of consciousness
remains inconclusive. Integrated information theory (IIT) argues that the posterior part of the brain is the hot
zone of consciousness, especially phenological consciousness. The IIT has proposed a “main complex”, a set of
elements determined such that the information loss in a hierarchical partition approach is the largest among
those of any other supersets and subsets, as the core of consciousness in a dynamic system. This approach may be
applicable not only to phenomenal but also to access-consciousness. This study estimated the main complex of
brain dynamics using functional magnetic resonance imaging in Human Connectome Project (HCP) and sleep
datasets. The complex analyses revealed the common networks across various tasks and rest-state in HCP,
composed of executive control, salience, and dorsal/ventral attention networks. The set of networks of the main
complex was maintained during sleep. However, compared with the wakefulness stage, the amount of information
of these networks and the default mode network, was reduced for the hypnagogic stage. The global
interconnected structure composed of major functional networks can comprise the core of consciousness.