train_node: Create hierarchical tree

In joseah/HierscPred: Supervised hierarchical classification for scPred

Description

Creates a tree object representing a cell-type hierarchy

Usage

train_node(
  tree,
  data,
  pvar,
  reduction,
  scaledata,
  model,
  allowParallel,
  verbose
)

Arguments

tree	Hierarchy created via create_hierarchy
data	Seurat object containing cells used to train hierarchical models
pvar	Prediction variable. Column name in Seurat object metadata containing the cell labels of the terminal nodes of the hierarchy tree
reduction	Name of reduction in Seurat objet to be used to determine the feature space. Default: "pca"
scaledata	Whether to select variable features and scale the data of the reference object. Default: FALSE
model	Classification model supported via caret package. A list of all models can be found here in https://topepo.github.io/caret/available-models.html
allowParallel	Allow parallel processing for resampling?
verbose	Print Seurat messages

Value

A data.tree object containing the cell type hierarchy

Author(s)

Lieke Michielsen and Jose Alquicira-Hernandez

Examples

branches <- c("Myeloid/DC/pDC", "Myeloid/DC/cDC")
h <- create_hierarchy(branches)
train_tree(data, h)

joseah/HierscPred documentation built on Nov. 25, 2021, 12:20 p.m.