STA 293B/BGT 08: Genomic Expression Analysis:
Technology, Computation & Analysis
K-means clustering routines
Acknowledgement: Some of the routines here are based on code from Roger Jang's
matlab site (among other things)
Data and Program files
- CellCycle.mat
Cell cycle data set (Matlab format)
- cellcycle_k.m
Example matlab code for running some K-means clustering analyses using cell cycle data
- kmeans.m
Does the shuffling of genes from cluster to cluster. This routine will
be modified in the near future to do exactly what was explained in class.
- show_clusters.m
Makes the plots of individual clusters.
- show_image.m
Plots a color map of clusters.
- show_genes.m
Retrieves data/descriptions pertaining to genes within a cluster.
CellCycle.mat
AD holds the data matrix
Des holds the descriptions
>> load CellCycle
>> whos
Name Size Bytes Class
AD 6519x14 730128 double array
Des 6519x48 625824 char array
V 1x17 34 char array
WLVERBOSE 1x3 6 char array (global)
ans 1x1302 2604 char array
Grand total is 405500 elements using 1358596 bytes
Examples of utilities
- show_clusters(AD,U) show plots of the clusters interactively
- show_image(AD,U,[1,2,3,4,5]) show an image of the clusters
1, 2, 3, 4, and 5
- show_genes(AD,U,3) return the rows of AD corresponding to
cluster 3
- show_genes(Des,U,3) return the Affymetrix identifiers of genes
in cluster 3