Table 1.

Summary of differences between Bulk RNA-seq and scRNA-seq.

	Goal	Protocol	Quality control	Normalization	Analyses
Bulk RNA-seq	• Measure the average gene expression across the population of cells in a sample • To identify differences between sample conditions	• RNA is extracted from all cells in the sample • Reverse transcription converts RNA to cDNA, facilitates ligation of sequencing adaptors • Amplification	• GC content, presence of adaptors, overrepresented k-mers, duplicated reads • Percentage of reads that map to reference • Reproducibility between replicates	• Batch effect • Between-sample variability: sequencing depth    Quantile normalization, spike-ins • Within-sample variability: feature length, library size effects    RPKM, FPKM, TPM	• Estimate gene and transcript expression • Differential expression analysis • Alternative splicing
scRNA-seq	• Measure the gene expression of individual cells in a sample • To identify differences between cell types/states	• RNA is extracted from isolated cells, labeled with cell specific identifier • UMIs, spike-ins often included, to account for higher levels of noise • Reverse transcription, amplification similar to bulk protocol	• Reads, number of genes per cell • Percentage of reads that map to spike-ins (if used), percentage of reads that map to mitochondria • QC metrics used in bulk RNA-seq are also examined	• Batch effect and within-sample variability are corrected for similarly to bulk RNA-seq • Between-sample variability methods must additionally account for capture efficiency and dropout sources of noise	• Dimensionality reduction • Identify cellsubpopulations • Differential expression • Pseudotime/trajectory analysis