Summary
International Symposium on Nonlinear Theory and Its Applications
2016
Session Number:C2L-E
Session:
Number:C2L-E-3
Large Graph Laplacian Matrix and Functional Map of Whole Brain of {\it C. elegans}
Hiromu Sakuma, Takayuki Teramoto, Sayuri Kuge, Takeshi Ishihara, Yuishi Iwasaki,
pp.-
Publication Date:2016/11/27
Online ISSN:2188-5079
Summary:
Towards understanding of neural signaling of Caenorhabditis elegans, cluster analysis is carried out for the whole-brain imaging data on the basis of spectral clustering. Correlation between the data is represented as a similarity graph in which each vertex corresponds to each data. Optimized clustering resolves into an eigenvalue problem of the graph Laplacian defined by the similarity graph. We analyze the neural activities of wild-type and unc-7 mutant which has defect in gap junction. We find that the neural activities of unc-7 mutant are divided into two large clusters and are weakly correlated with each other. For wild-type, on the other hand, correlation between the neural activities is relatively strong and the brain works as a whole.