I've made available some publicly available data containing single cell RNA-Seq data for 5k immune cells as well as a script that runs the Seurat workflow to define cell-type clusters in this data.
The script does the following:
- Load Data and Create Seurat Object
- Normalize Data Using SCTransform and Regress out Genes Related to Cell Cycle
- Identify Highly Variable Features
Scale the Data
- Select Dimensions using PCA (using variable features)
- Cluster the cells (using the selected number of dimensions)
- Find markers for each cluster
- Visualize the clusters using tSNE or UMAP
- Identify cells expressing genes of interest
- Label clusters as cell types based on expression of known marker genes.
You can get the data and R script from stampede2 here:
/work2/projects/BioITeam/projects/courses/rnaseq_course/day_5_single_cell_data
The script: seuratTotalScript5k.R
The data: filtered_feature_bc_matrix
Unfortunately, the packages required to run Seurat on TACC are either out-of-date or not installed. So, we are not going to run this script today.