Competitive interactions shape mammalian brain network dynamics and computation

Area of Science:

• Neuroscience
• Computational Biology
• Systems Neuroscience

Background:

• The brain's ability to integrate information relies on complex network architectures.
• Understanding how distributed circuits balance cooperative and competitive interactions is crucial for deciphering brain function.

Purpose of the Study:

• To investigate the dynamical and computational relevance of cooperative and competitive interactions within the mammalian connectome using whole-brain modeling.
• To determine how network architecture influences brain activity and computational performance.

Main Methods:

• Computational whole-brain modeling applied to human, macaque, and mouse connectomes.
• Analysis of emergent dynamical properties and subject-specific brain activity reproduction.

Main Results:

• Mammalian brain activity is best reproduced by models combining modular cooperation with diffuse, long-range competition.
• Competitive interactions preferentially link regions with opposing molecular and structural profiles.
• Models incorporating competition demonstrate superior subject specificity and fit to brain activity dynamics, with these properties emerging spontaneously.

Conclusions:

• A balance of cooperative and competitive interactions is fundamental to mammalian brain network architecture and function.
• Competitive interactions are key to achieving subject-specific brain dynamics and enhanced computational performance.
• This study establishes a generative link between network structure, dynamics, and computational capabilities in the brain.