Introduction to single-cell RNA-seq
Abel Vertesy
Disclaimer
This course material was heavily based on multiple sources:
- Kolodziejczyk, Aleksandra A., Jong Kyoung Kim, Valentine Svensson, John C. Marioni, and Sarah A. Teichmann. 2015. “The Technology and Biology of Single-Cell RNA Sequencing.” *Molecular Cell- 58 (4). Elsevier BV: 610–20. doi:[10.1016/j.molcel.2015.04.005](https://doi.org/10.1016/j.molcel.2015.04.005).
- https://github.com/hemberg-lab/scRNA.seq.course/
- Most figures are captured from the web & my apologies (to the authors) for missing references
History
Bulk RNA-seq
- Replaced microarrays in the late 00’s and has been widely used since
- Average expression level for 1000’s of genes
- Useful for:
- Compare tissues
- Compare disease / healthy
- Insufficient for studying
- heterogeneous systems
- early development
- complex tissues (brain)
- Stochastic gene expression
- heterogeneous systems
scRNA-seq
- A new technology, first publication by Tang 2009
- Widespread from ~2014: new protocols and lower sequencing costs
- Measures the distribution of expression levels for each gene
scRNA-seq evolution
Concepts
scRNA-seq allows to study new biological questions
- Cell type identification
- Heterogeneity of response
- Stochasticity of gene expression
- Reconstructing continous processeses from a single experiment
- Holistic view, hypothesis-free science
- Datasets range from 102 to 106 cells and increase in size every year
- Not testing a concrete hypothesis
- Hypothesis-driven and data-driven science
- Heard about it? Could you explain?
Hypothesis-driven and data-driven science
Experimental Strategies
Basic Workflow
from Wikipedia
Cell capture
Manual isolation by mouth pipet
Laser capture micro-dissection (LCM) into eppendorf tubes
FACS sorting into microwell plates
{width=600px}
Isolation by FACS machine
- Fluorescence activated cell sorting
- You can select a subset of cells:
- Size of blood cells
- Dazl-GFP mouse line to select germ cells
FACS concept
Isolation by microfluidics & preparation insider droplets
> Macosko 2015
Picowell
Picocells allow simple sedimentation instead of FACS
- Too small for FACS
- Volume per well is so low, that even at 10% loading, you waste very little reagents
Platforms & scaling
Source: Svensson 2018
RNA amplification strategies
- Full length
- 3’ (aka: 3 prime)
- 5’ (aka: 5 prime)
RNA structure
Expressed sequence tag counting (EST)
Different Coverage along the transcript (Full L. vs. 3’)
Bhargava 2014
Full length
- Advantage?
- Isoform specificity
- Higher sensitivity
- SMART-seq2 [@Picelli2013-sb]
3’
- Sufficient for transcript counting
- Advantage?
- A lot cheaper
- 3’ alternative polyadenylation
- One kind of isoforms
- CEL-seq (Hashimshony 2012)
- Drop-seq (Macosko 2015)
- Almost all novel methods
5’
- Sufficient for transcript counting & cheap
- 5’: Alternative transcription start site
- STRT-seq (Islam et al.)
Experimental Procedures in detail
Steps in single-cell mRNA sequencing
Problems
- RNA is unstable ← → RNAse enzymes are very robust
- Convert it to cDNA immediately
- Very low RNA input material → Huge amplification
- Huge amplification → Amplification bias (“PCR artefact”)
Reverse transcriptase
A reverse transcriptase (RT) is an enzyme used to generate complementary DNA (cDNA) from an RNA template, a process termed reverse transcription. Reverse transcriptases are used by:
- retroviruses to replicate their genomes,
- retrotransposon mobile genetic elements to proliferate within the host genome,
- eukaryotic cells to extend the [telomeres], and by
- some non-retroviruses such as the hepatitis B virus, a member of the Hepadnaviridae.
Wikipedia, edited
Reverse transcription creates uniquely tagged cDNA molecules
Velten 2015
Primer design & labelling strategy
| Sequence element | Function |
|---|---|
| poly-T stretch | Select poly-A RNA / deplete ribosomal RNA-s |
| UMI (Unique Molecular Identifier) | Label each mRNA molecule before amplification : correct for amplification bias at the end |
| CBC (Cell Barcode) | Each mRNA is labeled by cell of origin: samples can be pooled / multiplexed after this step → less labor |
| Illumina adapter | 2 of these are required for sequencing |
Unique Molecular Identifiers (UMIs),
Kivioja 2012
You can do better than just read counting, reads that map to the
- Same gene and have,
- same CBC &
- same UMI can be collapsed.
In-vitro transcription or PCR amplifies your
Velten 2015
Linearly amplification by IVT is less noise sensitive to fluctuation in input material
| PCR | IVT |
|---|---|
| aRNA = cDNA^n | aRNA = n * cDNA |
Kolodziejczyk 2015
- After the invention of UMI, it only spares sequencing costs
Information in paired end reads
Velten 2015
Overview
{width=300px}
Velten 2015
Introduction to single-cell RNA-seq
Abel Vertesy