bbknnR Tutorial
In bbknnR: Perform Batch Balanced KNN in R

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Setup library and data

library(bbknnR)
library(Seurat)
library(dplyr)
library(patchwork)
data("panc8_small")

Run BBKNN

Note that RunBBKNN() also compute t-SNE and UMAP by default.

panc8_small <- RunBBKNN(panc8_small, batch_key = "tech")

Find Clusters using bbknn graph

panc8_small <- FindClusters(panc8_small, graph.name = "bbknn")

Visualization

p1 <- DimPlot(panc8_small, reduction = "umap", group.by = "celltype", label = TRUE,
             label.size = 3 , repel = TRUE) + NoLegend()
p2 <- DimPlot(panc8_small, reduction = "umap", group.by = "tech")
p3 <- DimPlot(panc8_small, reduction = "umap")

wrap_plots(list(p1, p2, p3), ncol = 1)

p1 <- DimPlot(panc8_small, reduction = "tsne", group.by = "celltype", label = TRUE,
             label.size = 3 , repel = TRUE) + NoLegend()
p2 <- DimPlot(panc8_small, reduction = "tsne", group.by = "tech")
p3 <- DimPlot(panc8_small, reduction = "tsne")

wrap_plots(list(p1, p2, p3), ncol = 1)