Laplacian Eigenmaps for dimensionality reduction and data representation
Neural Computation
A kernel view of the dimensionality reduction of manifolds
ICML '04 Proceedings of the twenty-first international conference on Machine learning
A tutorial on spectral clustering
Statistics and Computing
IEEE/ACM Transactions on Computational Biology and Bioinformatics (TCBB)
Towards a theoretical foundation for Laplacian-based manifold methods
Journal of Computer and System Sciences
ISBRA'10 Proceedings of the 6th international conference on Bioinformatics Research and Applications
| Hi-index | 0.00 |
A range of experimental results indicate that the cell nucleus is highly organized, with many chromosomes forming relatively stable associations. An interesting prospect is that elements of this chromosomal organization may exist to facilitate coordinated gene expression. We present a flexible approach for detecting features of chromosomal domain organization that may be related to coordinated gene expression. The novelty of the presented approach is based on an application of nonlinear dimensionality reduction to organize genes with respect to a combined measure of co-expression and proximity along a chromosome. This allows identification of chromosomal neighborhoods over which genes are co-expressed. These locally correlated clusters yield a candidate expression-related inter-chromosomal interaction network with a prominent hub cluster. Our methods are demonstrated on a data set derived using 5 state-of-the-art gene expression profiling platforms over the widely-studied NCI-60 cancer cell lines. Two levels of network validation are presented: statistical, and with respect to experimentally measured physical interactions in a published study.