Skip to contents

Core Package Functions

Identifying marker genes, creating models, and classifying cells to a reference dataset.

FindMarkerGenes()
Finds marker genes at each hierarchical level specified by the tree, using the GE matrix provided in the ref_bpcells parameter, and the celltype labels provided in the input to ref_metadata. It identifies marker genes by using a function provided in the BPCells package, marker_features, which finds genes using the Wilcoxon test.
GetModels()
Function that gives models trained to differentiate all pairwise matchups.
Classify()
Classification function that takes in input dataset (in BPCells format), alongside created models and the tree structure and provides a vector of classifications.

Hierarchy Creation

To build a custom hierarchy.

CreateHierarchy()
Create Hierarchical Tree Structure from CSV with parent and child nodes named in a csv file
CreateEqualTree()
This function is meant for users who don't want to use the tree-based/hierarchical functionality of the package. It creates a tree that assumes that all the celltype labels are at the same level of the tree.

Example Datasets

test_ex_data_bpcells
PBMC 3k dataset subset
test_ex_metadata
PBMC 3k dataset testing metadata
train_ex_data_bpcells
PBMC 3k dataset subset
train_ex_metadata
PBMC 3k dataset training metadata