Identifying cell populations with scRNA-Seq

Single-cell RNASeq (scRNASeq) has emerged as a powerful method for quantifying the transcriptome of individual cells. However, the data from scRNASeq experiments is often both noisy and high dimensional, making the computational analysis non-trivial. Here researchers from the Karolinska Institutet provide an overview of different experimental protocols and the most popular methods for facilitating the computational analysis. They focus on approaches for identifying biologically important genes, projecting data into lower dimensions and clustering data into putative cell-populations. Finally they discuss approaches to validation and biological interpretation of the identified cell-types or cell-states.

Overview of methods covered in this review


Colour indicates which parts of the expression matrix are adjusted after each step, for instance feature selection only removes rows from the expression matrix, whereas dimensionality reduction calculates a new matrix composed of meta-features. Preprocessing steps not covered in detail in this review include quality control and normalization. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Andrews TS, Hemberg M. (2017) Identifying cell populations with scRNASeq. Mol Aspects Med [Epub ahead of print]. [abstract]

One comment

  1. We have been working on bulk RNAseq for gene expression for a while and I learned a lot from your blog. Recently I wanted to move onto scRNA-seq. We have a gene knockout mouse strain and suspect that some immune cells in their intestine might be different (in terms of gene expression, cell types, etc) from the wild-type mice. Since there are so many different types of intestinal immune cells, l wanted to use scRNA-seq to see if we could get some clues. The question is about the experimental design. We found some paper and discussion on the power analysis regarding how many cells to get enough power. However, I did not find how many biological repeats I should do. Say if I want to get ~5000 cells for scRNA-seq, should I collect intestinal immune cells from several mice and pour them together before loading it to 10x Genomics? Or I could just collect ~5000 cells from one or two mice? Your suggestion is appreciated.

Leave a Reply

Your email address will not be published. Required fields are marked *


Time limit is exhausted. Please reload CAPTCHA.