Edge-colored directed subgraph enumeration on the connectome.

Area of Science:

• Neuroscience
• Computational Biology
• Graph Theory

Background:

• Connectomics research generates complex neural wiring diagrams represented as graphs.
• Existing motif analysis often focuses on hypothesized structures, limiting discovery of novel biological functions.
• Current subgraph enumeration methods are computationally intensive and often ignore edge properties.

Purpose of the Study:

• To develop efficient computational strategies for large-scale subgraph enumeration in connectomics.
• To enable the discovery of novel, biologically significant motifs in neural networks.
• To account for different edge types (e.g., excitatory, inhibitory) in motif analysis.

Main Methods:

• A parallel, general-purpose subgraph enumeration strategy was developed to count motifs.
• A divide-and-conquer community-based approach was introduced for regional enumeration.
• The method incorporates differentiation of edge types to reflect biological properties.

Main Results:

• The proposed strategies were demonstrated on eleven connectomes.
• Extensive overviews of 26 trillion enumerated subgraphs were generated.
• The computation required approximately 9.25 years, highlighting the scale of the analysis.

Conclusions:

• The developed methods provide efficient tools for comprehensive motif discovery in large-scale neural connectomes.
• This work facilitates the identification of novel neural circuit motifs and their potential biological functions.
• The approach allows for detailed analysis of neural connectivity, considering diverse synaptic properties.