What is Single-Cell RNA-seq

The Problem with Bulk RNA-seq

Bulk RNA-seq measures the average gene expression across thousands or millions of cells. This is powerful for comparing conditions. But it hides an important reality: not all cells in a tissue are the same.

Imagine blending a fruit smoothie and trying to guess which fruits were in it. You can detect strawberry and banana flavors. But you cannot tell that one chunk was entirely strawberry while another was entirely banana. Bulk RNA-seq works the same way. It gives you the average across all cells, but it cannot tell you what individual cells are doing.

This matters because tissues contain many different cell types. A tumor sample contains cancer cells, immune cells, fibroblasts, and endothelial cells. Each cell type has a distinct gene expression profile. Bulk RNA-seq averages across all of them. If a gene is highly expressed in a rare cell population but silent everywhere else, bulk RNA-seq might show it as weakly expressed overall.

What Single-Cell RNA-seq Measures

Single-cell RNA-seq measures gene expression in individual cells. Instead of one measurement per sample, you get one measurement per cell. A typical experiment captures 5,000 to 20,000 cells. For each cell, you get a count of how many mRNA molecules were detected for each gene.

The result is a count matrix with genes as rows and cells as columns. Each entry tells you how many transcripts of that gene were detected in that cell. This matrix is the starting point for all downstream analysis.

With this data, you can:

Identify all cell types present in a tissue
Discover rare cell populations that bulk RNA-seq would miss
Track how cells change over time during development or disease
Compare cell type proportions between conditions

Key Discoveries Enabled by scRNA-seq

Single-cell RNA-seq has transformed our understanding of biology. Here are some landmark examples.

Rare cell types. In 2018, researchers used scRNA-seq to discover ionocytes in the lung airway. These rare cells express high levels of CFTR, the gene mutated in cystic fibrosis. Bulk RNA-seq had attributed CFTR expression to other cell types because ionocytes make up less than 1% of airway cells.

Tumor heterogeneity. Cancer is not a single disease within one patient. scRNA-seq reveals distinct subclones within a tumor. Some express drug resistance genes while others do not. This explains why some cancer cells survive treatment while others die.

Developmental trajectories. scRNA-seq can capture cells at different stages of differentiation. Computational methods reconstruct the path from stem cells to mature cell types. This has mapped developmental trajectories in the brain, blood, and gut.

Immune response mapping. During COVID-19, researchers used scRNA-seq to profile immune cells in infected patients. They identified which immune cell subtypes expanded, which became exhausted, and how the response differed between mild and severe disease.

Bulk vs Single-Cell RNA-seq

Feature	Bulk RNA-seq	Single-Cell RNA-seq
Resolution	Average across all cells	Individual cells
Cells per sample	Millions pooled together	500 to 50,000 profiled individually
Genes detected per sample	15,000 to 20,000	1,000 to 5,000 per cell
Cost per sample	$200 to $500	$2,000 to $10,000
Input material	100 ng to 1 ug RNA	Live single-cell suspension
Best for	Differential expression between conditions	Cell type discovery, heterogeneity
Replicates	3+ biological replicates standard	Often 1 to 3 samples per condition
Transcript coverage	Full-length or 3’ end	Usually 3’ end only

A key tradeoff is depth vs breadth. Bulk RNA-seq detects more genes per sample because it sequences deeply. Single-cell RNA-seq detects fewer genes per cell because the sequencing reads are spread across thousands of cells. This means lowly expressed genes are often undetected in individual cells. This phenomenon is called dropout.

How Single-Cell Sequencing Works

This video from Illumina provides an overview of the single-cell sequencing and analysis workflow, from sample preparation to data interpretation:

A Brief History

Single-cell RNA-seq has evolved rapidly since its invention.

2009: Tang et al. published the first scRNA-seq experiment, profiling a single mouse blastomere. The method was entirely manual: isolate one cell, lyse it, reverse transcribe, amplify, and sequence. Throughput was one cell at a time.

2012 to 2014: Plate-based methods like Smart-seq and Smart-seq2 improved sensitivity and captured full-length transcripts. Throughput increased to hundreds of cells per experiment, but each cell still required individual handling.

2015: Droplet-based methods changed everything. The Drop-seq and inDrop protocols encapsulated individual cells in nanoliter droplets, each containing a barcoded bead. This enabled profiling thousands of cells in a single experiment.

2016: 10x Genomics launched the Chromium platform, commercializing droplet-based scRNA-seq. The standardized kit made the technology accessible to any lab with the instrument.

2017 to 2020: The Human Cell Atlas project began cataloging every cell type in the human body. Datasets grew from thousands to millions of cells. Analysis tools like Seurat and Scanpy matured into full ecosystems.

2021 to present: Multimodal single-cell methods now measure RNA and protein simultaneously. Spatial transcriptomics adds location information. Single-cell ATAC-seq profiles chromatin accessibility. The field continues to expand beyond gene expression alone.

What Comes Next

The following pages in this section cover:

Experimental protocols: How the major platforms work, from 10x Genomics to Smart-seq
Analysis overview: The standard computational workflow from raw data to cell type annotations
Key challenges: Batch effects, annotation reproducibility, and why pseudobulk matters for differential expression