Skip to contents

Perform PCA-based feature or sample clustering using DynamicTreeCut.

Usage

auto_cluster_matrix_pca_one(
  mtx,
  mode = c("row", "col"),
  n_pcs = 20,
  method_hclust = "ward.D2",
  deepSplit = 0,
  minClusterSize = 10,
  pamStage = TRUE
)

Arguments

mtx

Numeric matrix. Rows or columns will be clustered.

mode

"row" or "col". Determines clustering direction.

n_pcs

Number of principal components to retain for distance calculation.

method_hclust

Agglomeration method for hierarchical clustering.

deepSplit

DynamicTreeCut parameter controlling cluster granularity.

minClusterSize

Minimum cluster size in DynamicTreeCut.

pamStage

Whether to apply PAM stage refinement in DynamicTreeCut.

Value

A data.frame containing `protein_group` and assigned cluster label.