Lecture 7 : Single Cell
RNA-seq analysis Vojta Bystry
vojtech.bystry@ceitec.muni.cz
Modern methods for genome analysis
(PřF:Bi7420)
RNA-seq types
● Bulk RNA-seq
● SC-RNA-seq
● Spatially resolved RNA-seq
Bulk RNA-seq vs. SC-RNA-seq
● Bulk RNA-seq
○ Several samples
○ Difference in RNA levels between predetermined set of samples (conditions)
● SC-RNA-seq
○ Hundreds to thousand cells
○ Distinguish (cluster) cells based on the difference in RNA levels
SC-RNA-seq primary analysis
● In principle similar to the bulk RNA-seq
○ Map to genomic reference, demultiplex and count reads per gene
● For 10x Genomics tool from the company - Cell Ranger
● Cell Ranger - report example
SC-RNA-seq primary analysis results
● per gene (feature) per cell read count
Pre-processing workflow
● Filtering of cells based on QC metrics
● Detection and filtering of highly variable features
● Data normalization and scaling
Filtering of cells QC
● Number of genes
● Number of reads
● % of mitochondrial RNA
○ Sign of dying cells
● Example filtering:
○ filter cells that have unique feature counts over 2,500 or less than 200
○ filter cells that have >5% mitochondrial counts
Identification of highly variable features
Determine the ‘dimensionality’ of the dataset
● How many clusters (cell types) is in the experiment (we want
to explore)
● Analysis of PCA results
● Better to choose higher value then smaller
Cluster the cells
● graph-based algorithms
● UMAP/tSNE visualization
Analysing clusters
● Finding differentially expressed features (cluster biomarkers)
Assigning cell type identity to clusters
Advanced analysis
● Cell state/development directions
○ Pseudo-time
○ Velocity
Pseudotime
● does not provide information about directionality of dynamics
Velocity
16www.ceitec.eu
CEITEC
@CEITEC_Brno
Vojta Bystry
vojtech.bystry@ceitec.muni.cz
Thank you for your attention!