Overview
Single-cell RNA sequencing (scRNA-seq) measures the transcriptome of individual cells, providing unprecedented resolution of cellular heterogeneity. Bulk RNA-seq averages gene expression across millions of cells, masking the distinct transcriptional states of different cell types. In contrast, scRNA-seq captures the expression profile of each cell separately, enabling the discovery of rare cell populations, the reconstruction of developmental trajectories, and the characterization of cell-type-specific responses. Since the first scRNA-seq study in 2009, the technology has advanced rapidly, with modern platforms profiling tens of thousands of cells in a single experiment.
Methods
The scRNA-seq workflow involves cell isolation (using droplet-based platforms such as 10x Genomics Chromium, microfluidics like Fluidigm C1, or plate-based methods), reverse transcription with unique molecular identifiers (UMIs) to remove amplification bias, and library preparation for sequencing. Bioinformatics analysis proceeds through: quality control (filtering cells by gene count, mitochondrial content, and UMI counts), normalization (SCTransform, scran, or BASiCS), batch correction (Harmony, Seurat’s CCA, or scVI), dimensionality reduction (PCA followed by UMAP or t-SNE visualization), clustering (Louvain or Leiden algorithms), and differential expression between clusters. Cell type annotation uses marker genes and reference databases.
Applications
scRNA-seq has transformed multiple fields. It has created comprehensive cell atlases for human organs, mapped the developing embryo, and revealed previously unknown cell subtypes in the brain and immune system. In cancer research, scRNA-seq dissects tumor heterogeneity and the tumor microenvironment, identifying rare drug-resistant cell states related to flow cytometry markers. The technique builds on core RNA sequencing principles and DNA microarrays and gene expression concepts. Emerging spatial transcriptomics methods now add positional context to single-cell data, revealing how cells organize within tissues.
Practical Protocol
A typical 10x Genomics scRNA-seq experiment begins with preparing a single-cell suspension at a concentration of 700–1200 cells/µL with viability above 90%. The Chromium Controller partitions cells into nanoliter-scale Gel Bead-In-Emulsions (GEMs), each containing a barcoded bead with unique 10x barcodes and UMIs. After reverse transcription within GEMs, the barcoded cDNA is pooled and amplified. Library construction involves enzymatic fragmentation, end repair, A-tailing, adapter ligation, and sample index PCR. Sequencing is performed on an Illumina platform with a read configuration of R1 28 cycles (10x barcode and UMI), R2 91 cycles (cDNA insert), and i7 index 8 cycles. The Cell Ranger pipeline processes raw BCL files into FASTQs using cellranger mkfastq, then aligns reads to the reference genome with STAR via cellranger count. The output includes filtered feature-barcode matrices, UMAP embeddings, and clustering results. Downstream analysis in R uses Seurat: CreateSeuratObject, SCTransform for normalization, RunPCA and RunUMAP for dimensionality reduction, and FindClusters with the Leiden algorithm. In Python, Scanpy provides analogous functions: sc.pp.filter_cells, sc.pp.normalize_total, sc.pp.log1p, sc.tl.pca, and sc.tl.leiden. Doublet detection with DoubletFinder or scrublet is recommended before clustering, and batch integration across samples can be performed with Harmony or scVI.
