scRNAseq: configure UMAP dimension reduction in Seurat R-package #UMAP #scRNAseq #Seurat

UMAP dimension reduction in Seurat R-package used umap-learn python library.

To configure this analysis is important to make interact these two languages : python and R

#in shell install pip sudo apt install python-pip pip install umap-learn pip list #Or install umap-learn in R via reticulate install.package("reticulate") library(reticulate) reticulate::py_install(packages ='umap-learn') #in shell setup your python virtualenv /usr/bin/python2.7 -m pip install --upgrade --user virtualenv #return in R library(Seurat)

#on Seurat objet post running PCA pbmc <- RunUMAP(pbmc, dims = 1:5) library(ggplot2) pdf("umap.pdf") DimPlot(pbmc, reduction = "umap", pt.size = 1) + ggtitle(label = "UMAP") dev.off