Constructs a phylogenetic tree relating the 'average' cell from each identity class. Tree is estimated based on a distance matrix constructed in either gene expression space or PCA space.
BuildClusterTree(object, genes.use = NULL, pcs.use = NULL, SNN.use = NULL,
do.plot = TRUE, do.reorder = FALSE, reorder.numeric = FALSE)Seurat object
Genes to use for the analysis. Default is the set of variable genes (object@var.genes). Assumes pcs.use=NULL (tree calculated in gene expression space)
If set, tree is calculated in PCA space, using the eigenvalue-WeightedEucleideanDist distance across these PC scores.
If SNN is passed, build tree based on SNN graph connectivity between clusters
Plot the resulting phylogenetic tree
Re-order identity classes (factor ordering), according to position on the tree. This groups similar classes together which can be helpful, for example, when drawing violin plots.
Re-order identity classes according to position on the tree, assigning a numeric value ('1' is the leftmost node)
A Seurat object where the cluster tree is stored in object@cluster.tree[[1]]
Note that the tree is calculated for an 'average' cell, so gene expression or PC scores are averaged across all cells in an identity class before the tree is constructed.