The braingraph.org database of high resolution structural connectomes and the brain graph tools

Area of Science:

• Neuroscience
• Computational Biology
• Graph Theory

Background:

• Understanding human brain connectivity is crucial for neuroscience.
• Existing datasets often lack detailed structural connectome information at various resolutions.
• Standardized data formats are needed for efficient brain data processing and visualization.

Purpose of the Study:

• To compute and provide structural connectomes for a large cohort of human subjects.
• To offer brain graph data at multiple resolutions and with various edge weights.
• To make brain connectome data accessible in a standardized, anatomically annotated format.

Main Methods:

• Utilized data from the Human Connectome Project for 426 subjects.
• Computed structural connectomes at five resolutions (83, 129, 234, 463, 1015 nodes).
• Generated anatomically annotated GraphML files for processed connectomes.
• Extracted and provided anatomically classified sub-graphs for 96 subjects.
• Included partially directed connectomes for 423 subjects.
• Developed and released the Brain Graph Tools for data processing.

Main Results:

• Structural connectomes computed for 426 human subjects across five resolutions.
• Anatomically annotated GraphML data available, facilitating downstream analysis.
• Accessible sub-graphs for specific brain regions (e.g., frontal lobes, precuneus) for 96 subjects.
• Partially directed connectomes available for 423 subjects.
• Brain Graph Tools software suite released on GitHub for connectome processing.

Conclusions:

• The study provides a comprehensive, multi-resolution dataset of human structural connectomes.
• The standardized GraphML format and available tools enhance usability for researchers.
• This resource facilitates advanced studies in brain connectivity, function, and disorders.