Clustering Coefficient

Currently

What we know

1. Autism spectrum disorder is characterized by altered network topology, featuring reduced functional connectivity within core networks like the default mode, visual, and sensorimotor systems, and less distinct boundaries between these networks.
2. Atypical structural and functional brain network properties, including altered white matter tract integrity and atypical developmental efficiency trajectories, correlate with the severity of social and communication symptoms in autism.
3. Down Syndrome is associated with a less specialized network structure marked by abnormally high global synchrony across major brain networks, which inversely correlates with performance IQ.
4. Despite exhibiting high internal brain synchrony, individuals with Down Syndrome show reduced synchronization with other viewers when processing external naturalistic stimuli, indicating highly individualized processing.
5. In Alzheimer's disease, tau pathology targets highly connected hub regions and degrades cortical connection strength, whereas in progressive supranuclear palsy, tau accumulates in deep structures and forces the cortex to use less efficient, indirect pathways.

Procedure

How it works

The main stages of the method. Specific protocols and analysis choices vary between studies.

1. 1

   Build a brain graph

   Regions become nodes and a structural or functional association becomes an edge, usually after thresholding a connectivity matrix. The threshold and whether the graph is weighted or binary decide how many triangles can exist; they are analysis choices, not properties of the brain.

2. 2

   Count closed neighbourhoods

   For each node the clustering coefficient is the fraction of that node's neighbors that are also connected to each other. High local clustering is a statement about triangles in the graph, often read as segregation, not a microscope on specialized tissue.

3. 3

   Summarise locally or globally

   Nodal values can be mapped, or averaged into a global clustering coefficient, sometimes compared with random or lattice null graphs. Global summaries wash out regional differences that may be the scientifically relevant part.

4. 4

   Do not equate clustering with better cognition

   The metric can rise when a network is pathologically synchronised and poorly segregated. A higher clustering coefficient is not automatically a marker of superior function, and it does not by itself explain a clinical score.

Common misconceptions

• ✗A higher clustering coefficient or increased synchronization across the brain is always a marker of superior cognitive functioning.

  ✓Abnormally high global synchronization can indicate a less specialized, poorly segregated network, as seen in Down Syndrome where elevated synchrony correlates with lower performance IQ.

• ✗Brain network architecture and structural efficiency trajectories remain constant during adolescent development in individuals with neurodevelopmental disorders.

  ✓Adolescents with autism show altered or absent typical developmental trajectories toward globally efficient structural networks over time.

• ✗Tau pathology impairs functional connectivity uniformly across different tauopathies.

  ✓The impact is disease-specific: Alzheimer's tau targets cortical hubs and weakens cortical connectivity, while progressive supranuclear palsy tau accumulates in deep structures, forcing reliance on indirect cortical pathways and increasing cortical-to-cortical connectivity.

Related

• Functional Connectivity
• Resting-State Functional Magnetic Resonance Imaging
• Diffusion Tensor Imaging
• Resting-State Functional Connectivity
• BOLD Signal