Clustering methods are used to group similarly expressed genes or experiments into subsets, or clusters, in ArrayStar’s Heat Map view.
Two powerful clustering algorithms are available: Hierarchical Clustering (Clustering > Hierarchical) and k-Means Clustering (Clustering > k-Means).
If desired, you can edit clustering method options using the Clustering Parameters dialog (Clustering>Advanced Clustering).
Note that clustering is only available for gene data and cannot be used to group IP Peak or IP Fragment data. Furthermore, genes must have binding activity in at least some of their conditions in order to be clustered. In a ChIP-Seq project, most of the genes do not have binding activity near them, so clustering such genes is not a valid operation. To make an appropriate ChIP-Seq gene selection, you may wish to try selecting genes from the Scatter Plot rather than the Gene Table.
Note: If your experiment involves data normalized using DESeq2 or edgeR, clustering uses the rlog values (log2) for each gene