- Curvature Motifs in Connectome Data: Pitfalls with Path Motifs
- Project Year:
2022
- REU Student (s):
Liron Karpati | University of Maryland-College Park MD
- Student 1 Institution:
University of Maryland-College Park
- Project Mentor:
Jie Gao
- Project Mentor Area:
Computer Science
- Project Abstract:
Nervous systems are organized for efficient integration of information. It has been shown that, in neuronal connectomes of the C. Elegans worm, high degree nodes are highly connected to form what is called a "rich-club" structure. This rich-club structure allows different neurons to reach one another in relatively few edge hops. The rich-club organization is a global architectural feature of the C. Elegans connectome. It has yet to be explored how the local organization of the connectome is supporting integration integration. By using the notion of course Ricci curvature (a generalization of Ricci curvature) and a path motif analysis, we found that the C. Elegans connectome exhibits local weak-bridge structures. More importantly, our analysis highlights some of the critical pitfalls in hypothesis testing graph properties. These pitfalls motivate the need for a principled investigation into statistical methods for determining the significance of graph properties.
