No lecture slides for this class. We’ll spend the entire time working on scRNA-seq script.

Overview

Last week, you learned the basics of single cell data analysis, including QC and filtering, clustering, integration of mutliple samples, and cluster identification using reference databases. To build on this, in the final lecture of the course, we’ll delve into more advanced topics for single cell analyses. By the end of this lecture, you’ll be able to handle multi-modal data, carry out subclustering of your favorite cell types, and use ‘pseudobulking’ for improved differential expression analysis…and more!

Learning objectives

  • Improve comfort with handling Seurat objects and understanding assays, idents, and metadata.
  • Practice producing and interpreting QC analysis of single cell data
  • Learn how to customize UMAP plots for publication
  • Use Weighted Nearest Neighbor analysis to combine data from different modalities
  • Learn about cell cycle scoring and regression
  • Gain experience focusing in on a specific cell type using subclustering
  • Learn to use ‘pseudobulking’ to improve differential expression analysis

Code and files

DIY_scRNAseq_advanced.R - this is the R script that we’ll use for this lecture.

Seurat object - You’ll need this Seurat object to follow along with the script above in the video lectures below. The object contains data from messenteric lymph node (MLN) cells collected from mice harboring different kinds of intestinal infections. MLNs are a central ‘hub’ where immune cells from the gut congregate to interact and mount a robust immune response.

Lecture videos

Video production is currently in progress. Check back soon!

Reading

Integrated analysis of multimodal single-cell data - This paper from the Rahul Satija’s group (creater of the Seurat package) describes the unsupervised weighted nearest neighbor approach for integrating multimodal data (e.g. RNA-seq and ATAC-seq) in single cell studies.

Inference and analysis of cell-cell communication using CellChat – as the title suggests, this paper outlines the use of CellChat for inferring cell-cell interactions from single cell RNA-seq data by examining receptor-ligand expression patterns.

Single-cell RNA-seq reveals changes in cell cycle and differentiation programs upon aging of hematopoietic stem cells – This publication is the source of the cell cycle genes that we use to assess and control for cell cycle in our UMAP analysis.