Glossary

scRNA-seq

Single-cell RNA sequencing, a high-throughput technology that profiles the transcriptome of individual cells to resolve heterogeneity masked in bulk sequencing.

Get in touch Learn more

Moody home-office setup in a converted highrise loft, analyst working late with multiple screens showing knowledge graph visualizations, city lights through large windows behind.

SINGLE-CELL TRANSCRIPTOMICS

What is scRNA-seq?

Single-cell RNA sequencing (scRNA-seq) is a high-throughput technology that profiles the transcriptome of individual cells, resolving cellular heterogeneity that is masked in traditional bulk RNA sequencing. By capturing and sequencing the messenger RNA (mRNA) from thousands of single cells in parallel, scRNA-seq enables the discovery of novel cell types, states, and dynamic biological processes within complex tissues.

Single-cell RNA sequencing (scRNA-seq) is a high-throughput technology that profiles the entire transcriptome—the complete set of messenger RNA (mRNA) transcripts—within individual cells. Unlike bulk RNA-seq, which averages gene expression signals across a population of cells, scRNA-seq resolves cellular heterogeneity by capturing the unique transcriptional signature of each cell. This enables the identification of rare subpopulations, transient cell states, and continuous developmental trajectories that would otherwise be obscured.

The core workflow involves isolating single cells into droplets or microwells, lysing them, and capturing their mRNA using barcoded primers. Each transcript is tagged with a Unique Molecular Identifier (UMI) and a cell-specific barcode before reverse transcription, amplification, and sequencing. The resulting count matrix—where rows represent genes and columns represent individual cell barcodes—serves as the foundation for downstream computational analysis, including dimensionality reduction, graph-based clustering, and trajectory inference.

SINGLE-CELL RESOLUTION

Key Features of scRNA-seq

Single-cell RNA sequencing resolves transcriptomic heterogeneity by profiling individual cells rather than bulk populations. These core features define the technology's analytical power and computational requirements.

Transcriptome-Wide Profiling at Single-Cell Resolution

scRNA-seq captures the entire polyadenylated transcriptome of individual cells, quantifying expression levels for thousands of genes simultaneously. Unlike bulk RNA-seq, which averages signals across heterogeneous populations, this approach reveals rare cell types, transient states, and stochastic gene expression patterns. The resulting count matrix typically contains 10,000–30,000 genes measured across 1,000–1,000,000+ cells, depending on the protocol.

Unique Molecular Identifier (UMI) Integration

Modern droplet-based and plate-based protocols incorporate Unique Molecular Identifiers—random 8–12 nucleotide barcodes—during reverse transcription. Each transcript molecule receives a distinct UMI, enabling absolute molecular counting rather than relative abundance estimation. This eliminates PCR amplification bias and allows computational removal of duplicate reads, producing a more accurate representation of the original transcript population.

High-Dimensional Feature Space

Each cell is represented as a vector in a high-dimensional gene expression space where dimensions equal the number of detected genes. This richness enables fine-grained discrimination of cellular identities but introduces the curse of dimensionality—distances become less meaningful, and computational costs escalate. Dimensionality reduction techniques like PCA, t-SNE, and UMAP are essential for visualization and noise filtering.

Sparse Count Data with Dropout Events

scRNA-seq data is characterized by zero-inflated distributions where a large fraction of genes show zero counts in any given cell. These zeros arise from both biological absence (the gene is truly not expressed) and technical dropout (the transcript was present but not captured during reverse transcription). This sparsity—often exceeding 90% zeros—requires specialized statistical models that account for the bimodal nature of the data.

Cell Barcoding and Multiplexing

Each cell receives a unique oligonucleotide barcode during library preparation, enabling thousands of cells to be sequenced simultaneously in a single run. Advanced multiplexing strategies like cell hashing use antibody-conjugated barcodes to label cells from different samples, allowing computational demultiplexing after sequencing. This dramatically reduces per-sample costs and eliminates batch effects between samples processed in the same lane.

Multimodal Readout Capabilities

Beyond transcriptome profiling, extended protocols like CITE-seq simultaneously measure surface protein abundance using oligonucleotide-conjugated antibodies. scATAC-seq captures chromatin accessibility, while spatial transcriptomics preserves tissue location information. These multimodal assays generate paired measurements from the same cell, enabling integrated analysis of gene expression, protein levels, and epigenetic states within a unified computational framework.

SINGLE-CELL RNA SEQUENCING

Frequently Asked Questions

Clear, technical answers to the most common questions about single-cell transcriptomics, from fundamental mechanisms to computational analysis.

Single-cell RNA sequencing (scRNA-seq) is a high-throughput technology that profiles the entire transcriptome of individual cells, resolving cellular heterogeneity that is masked in bulk RNA sequencing. The process begins with single-cell isolation using techniques like droplet-based microfluidics (10x Genomics), microwell-based capture, or fluorescence-activated cell sorting (FACS). Each isolated cell is lysed, and its mRNA is captured, reverse-transcribed into cDNA, and amplified. A critical innovation is the incorporation of Unique Molecular Identifiers (UMIs) —random barcodes that tag individual transcripts before amplification—enabling absolute molecular counting and computational removal of PCR duplicates. The resulting libraries are sequenced on platforms like Illumina NovaSeq, producing reads that are aligned to a reference genome and quantified into a count matrix, where rows represent genes and columns represent individual cell barcodes. This matrix serves as the foundation for all downstream computational analysis, including dimensionality reduction, clustering, and differential expression testing.

Enabling Efficiency, Speed & Accuracy

Intelligent Analysis, Decision & Execution

We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.

Talk to Us

Search across company data

Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.

Useful when people spend too long searching or get different answers from different systems.

Enterprise searchRAGPermissions

Automate internal workflows

Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.

Useful when repetitive work moves across multiple tools and teams.

AI agentsWorkflow automationGovernance

Add AI to products and internal tools

Build assistants, guided actions, or decision support into the software your team or customers already use.

Useful when AI needs to be part of the product, not a separate tool.

AI integrationDecision supportModel routing

EXPLORE THE ECOSYSTEM

Related Terms

Mastering scRNA-seq requires understanding the computational and experimental methods that surround it. These core concepts form the analytical backbone of single-cell transcriptomics.

Count Matrix

The fundamental data structure of scRNA-seq. A sparse numerical matrix where rows represent genes and columns represent individual cell barcodes. Each entry stores the number of unique transcripts detected for a specific gene in a specific cell. Due to the high dimensionality and sparsity (mostly zeros), specialized data formats like AnnData or loom are used for efficient storage and computation.

>90%

Typical Sparsity

Dimensionality Reduction

Mathematical transformation of high-dimensional single-cell data into a lower-dimensional space for visualization and noise reduction. Common techniques include:

PCA: Linear method capturing global variance.
t-SNE: Non-linear method preserving local neighborhood structure.
UMAP: Graph-based method balancing local and global structure, often preferred for speed and cluster preservation.

Graph-Based Clustering

An unsupervised method that partitions cells into groups by constructing a k-nearest neighbor (KNN) graph in a reduced dimensionality space. Community detection algorithms like Louvain or Leiden are then applied to identify densely connected communities. The resolution parameter controls the granularity of the resulting clusters, directly impacting cell type identification.

RNA Velocity

A computational method that predicts the future transcriptional state of individual cells by distinguishing between unspliced (nascent) and spliced (mature) mRNA reads. The ratio of these molecules provides a directional vector, allowing researchers to infer developmental trajectories and cellular transitions without requiring time-series experiments.

Batch Effect Correction

Non-biological systematic variation introduced by technical factors like different experimental runs, reagents, or sequencing lanes. Data integration methods such as Harmony, scVI, or Seurat's CCA align multiple datasets to remove these technical artifacts while preserving true biological variation, enabling valid cross-condition comparisons.

Cell Type Annotation

The assignment of biological identity labels to cell clusters. This is performed either manually using curated marker gene databases or automatically via label transfer from a well-characterized reference atlas. Automated classifiers use shared latent space representations to project labels onto query datasets, standardizing nomenclature across studies.

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.

Limited slotsGet a Free AI Consultation

How We Work

Custom AI workflows for your Business

One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.