Order of samples and variables in hierarchical clustering

From Array Suite Wiki

Jump to: navigation, search


In the resulting heatmap (or dendrogram) of hierarchical clustering, there is no "quick" way to reorder samples or variables if both variable tree and observation tree have been computed. If user wants to retain certain orders of the samples and variables in the heatmap, the raw data should be reordered first and rerun of the hierarchical clustering is needed.

Contents

Reorder the raw data

User can generate a list of variables (samples) of interest in desired order, and then use "Reorder rows by list" to reorder the raw data. Choose "Annotation" to order variables and "Design" to order samples.

20131104 OrderHierarchicalClustering 2.png

Reorder the raw data by factor level

If the user just wants to reorder sample by the levels of certain experiment factor (e.g. Treatment levels A, B and C), he/she can use Reorder Factor Levels function.


Reorder the raw data by Subset

The Subset function can be used to reorder variables and samples. see more details in Reorder variables or observations in Microarray data.


Retain order by customized list

User can generate a list of variables (samples) of interest in desired order, and then provide the list when specifying data input in the hierarchical clustering analysis. This way will not change the order in the raw data.

20131104 OrderHierarchicalClustering 1.png