Custom Analysis         Clustering Example Experiment: Sample A and Sample B are compared at six time points, and fold-changes (or logRatios) are calculated for the six time points. The expression profile for each gene describes how the (relative) expression level changes over time. Genes with similar profiles are grouped together. There are various ways of calculating the “similarity” among genes, e.g. hierarchical clustering, K-mean clustering, etc. (1) Hierarchical clustering: (2) K-mean clustering Expression levels are color coded: green being down-regulated and red up-regulated. Six columns correspond to six time points, and each row represents the expression profile for over time for one gene. Genes with similar expression profiles are grouped closely on the heatmap. Similarly samples (i.e. different time points) can be grouped according to the gene expression levels, as shown by the dendrogram on the top of the figure.   Genes are partitioned into several groups and as shown in the figure, genes in the same group have similar expression patterns over time; while genes in different groups have different expression patterns. The histogram shows the number of genes in each group. K means clustering (Page 1) K means clustering   << Back to Custom Analysis         Copyrights © 2006 MOgene, LC