docs/3plaintext.md

In SGDDNB/ShinyCell: Shiny Interactive Web Apps for Single-Cell Data

title: | | Tutorial on processing plain-text gene expression matrices author: | | John F. Ouyang date: "Feb 2021" output: pdf_document: default html_document: toc: true toc_depth: 2 toc_float: collapsed: false fontsize: 12pt pagetitle: "3plaintext"

Introduction

Apart from major single-cell data formats (h5ad / loom / Seurat / SCE), users may start with a plain-text gene expression matrix or output from the 10X cellranger pipeline. Here, we provide a simple tutorial on performing some basic single-cell analysis using the Seurat pipeline and creating a ShinyCell app using the generated Seurat object.

WARNING: We would like to emphasize that ShinyCell is a visualisation tool and not an automated analysis tool. So while we provide some code to perform basic single-cell analysis, it is up to the user to ensure that the Seurat analysis makes sense and agree with their own biological understanding.

Code

The code comprises several parts. First, after loading the required packages, we will read in the plain-text gene expression matrix and convert it into a sparse matrix format. For cellranger outputs, one can run the Read10X() instead to read in the gene expression from the cellranger output directory. Second, we will perform some basic single-cell analysis using the Seurat pipeline. The analysis include preprocessing the data, followed by PCA and UMAP dimension reduction and then unsupervised clustering. Third, we will create a ShinyCell app using the generated Seurat object. The code is essentially the same as that in the quick start guide.

library(data.table)
library(Matrix)
library(Seurat)
library(ShinyCell)


## 1. Read in gene expression
getExampleData("plaintext")  # Download plain text example dataset
inpGEX = fread("test/rset.txt.gz")
inpGEX = as.matrix(inpGEX[, -1], rownames = inpGEX$V1)
inpGEX = as(inpGEX, "dgCMatrix")
# inpGEX = Read10X(data.dir = cellranger/output/directory/)


## 2. Seurat pipeline
seu = CreateSeuratObject(counts = inpGEX)
seu = NormalizeData(seu)
seu = FindVariableFeatures(seu, nfeatures = 1500)
seu = ScaleData(seu)

seu = RunPCA(seu, npcs = 30)
seu = RunUMAP(seu, dims = 1:15)
seu = FindNeighbors(seu, dims = 1:15)
seu = FindClusters(seu, resolution = 0.5, random.seed = 42)

## 3. ShinyCell
scConf = createConfig(seu)
makeShinyApp(seu, scConf, gene.mapping = TRUE, 
             shiny.dir = "shinyAppPlain",
             shiny.title = "ShinyCell Plain Text Tutorial") 

Running the above code generates a shiny app in the shinyAppPlain/ folder, looking like this:

SGDDNB/ShinyCell documentation built on Jan. 25, 2024, 3:19 p.m.