Perturb-Seq
Perturb-Seq

Perturb-Seq

Bridging high-throughput CRISPR perturbations and single-cell RNA sequencing to unlock functional genomic insights

What is Perturb-Seq?

Perturb-Seq is a high-throughput method that combines the precision of clustered regularly interspaced short palindromic repeats (CRISPR) perturbations with single-cell RNA sequencing (scRNA-Seq). This combined approach is used to study the transcriptome-wide effects of genetic perturbations at single-cell resolution, investigate gene regulatory networks (GRNs) and cellular responses, and provide insights into human health and disease.1

Historically, the cost of scRNA-Seq and workflow constraints have limited the scale and throughput of studies that use Perturb-Seq. These bottlenecks are problematic for studies that seek to map GRNs, which provide a mechanistic link between genetic variation and human traits. While single-cell transcriptomics offers a high-resolution readout of transcriptional states, it is difficult to use observational data to infer causal regulatory relationships.2,3

Genome-scale Perturb-Seq screens can help address these limitations by introducing targeted genetic perturbations, enabling researchers to identify gene function, establish causal regulatory relationships, and gain functional genomic insights. To help overcome historical throughput and workflow barriers, Illumina offers next-generation sequencing (NGS) solutions that enable high-throughput Perturb-Seq experiments.

How Perturb-Seq works

High-throughput Perturb-Seq experiments follow a workflow that uses a CRISPR-based screen to simultaneously disrupt genes from a focused handful to several thousand within a population of cells. Common CRISPR screens include CRISPR interference, CRISPR-Cas9 for knockouts, and CRISPR activation. Following the CRISPR-based screen, cells are further processed for single-cell library preparation.3 Libraries are subsequently sequenced on high-throughput sequencing systems, such as the NovaSeq X Series, and the data is analyzed using single-cell pipelines.

Profile clip of a female scientist placing a library tube strip into a reagent cartridge to prepare for a NovaSeq X run; blurry instrument in foreground.

Perturb-Seq vs traditional pooled CRISPR screens

While traditional pooled CRISPR screens are excellent when investigating one-dimensional phenotypes, including cell survival and fitness readouts, Perturb-Seq captures the whole-genome expression profile of individual cells. The high-dimensional transcriptomic signature obtained from Perturb-Seq experiments offers high phenotypic resolution and a window into mapping gene regulatory networks.3,4

Feature Traditional pooled CRISPR screens Perturb-Seq
Phenotypic readout Low-dimensional readout, provides cell viability (survival), proliferation (fitness), or fluorescent-activated cell sorting (FACS) information based on a small selection of predefined markers High-dimensional readout, scRNA-Seq expands the scope of traditional pooled CRISPR screens by capturing the whole-transcriptome signature
Data resolution Average of the bulk cell population, masking cell-to-cell heterogeneity Single-cell resolution, providing detailed profiling of transcriptional states
Primary output metric Single guide RNA (sgRNA) enrichment or depletion (eg viability or fitness scores) Transcriptomic shifts per each perturbation and differential gene expression
Primary applications Examples include discovering primary cell survival pathways, essential genes, and drug resistance mechanisms Examples include mapping GRNs, identifying potential drug targets, and uncovering epistatic interactions at the transcriptome level
Scale and throughput profile Scalable to large genome-wide screens (~50–100K sgRNAs) in a single bulk run Historically targeted, but emerging multiomics and probe-based workflows are scaling to genome-wide capacity

Illumina Perturb-Seq solutions

Illumina Single Cell 3' RNA Prep

Scalable RNA Prep kits that offer high transcript and gene sensitivity for single-cell experiments without complex workflows or microfluidics.

NovaSeq X Series

NovaSeq X Series ultra-high-throughput, production-scale sequencing systems enable more accessible and efficient genome-wide Perturb-Seq experiments.

DRAGEN Single Cell RNA analysis

This workflow optimizes single-cell sequencing analysis and is suitable for use with the Illumina Single Cell 3'RNA Prep while also supporting third-party single cell preps.

Unlock the next wave of genomic discovery

Learn about studies that were once unimaginable with advancements in high-throughput genomics, multiomic studies, and genome-scale Perturb-Seq to help dissect complex cellular pathways, and how the NovaSeq X Sequencing Systems are accelerating genomic insights.

Genetic perturbation data set for drug discovery

Through collaborations with pharmaceutical partners using AI and human genomics, Illumina is assembling the Billion Cell atlas to be the world’s largest genome-wide genetic perturbation data set. Learn how this immense resource will cover numerous cell types and enable researchers to better understand perturbation prediction, annotation, cell-state, and drug-response modeling. Find out how to access this breakthrough data set.

Learn more about the Illumina Billion Cell Atlas
Illumina introduces Billion Cell Atlas press release
Thumbnail

Scaling Perturb-Seq with Illumina Single Cell Prep

Cost and workflow restrictions have historically made scRNA-Seq difficult to perform at scale. Hear from Sheila Dodge, Chief Operating Officer at Broad Clinical Labs, as she shares insights on scaling Perturb-Seq to profile five million cells in five days.

Perturb-Seq FAQ

Both are single-cell CRISPR technologies that link genetic perturbations to a transcriptomic readout. The primary difference between CRISPR droplet sequencing (CROP-Seq) and Perturb-Seq resides in vector design and how the guide RNA is captured and identified during sequencing.

Early Perturb-Seq vectors included an indirect barcode for sequencing and were limited due to lentiviral recombination errors that could occur, uncoupling the barcode from the guide. CROP-Seq solved this issue by linking the guide to the transcriptome, circumventing possible recombination errors. Today, the original indirect capture methods are obsolete as modern Perturb-Seq assays largely use direct-capture vectors.5

Learn more about NGS in CRISPR genome editing and the advantages of scRNA-Seq.

Traditionally, droplet-based, single-cell workflows and relevant CRISPR-compatible single-cell RNA prep kits have provided scientists with Perturb-Seq options. Both the cost of scRNA-Seq and workflow limitations constrain the throughput and scale of these experiments.6 Additionally, platform suitability depends on several factors, including scale of the screen, read depth requirements, and cost constraints among other considerations.

Powering scRNA-Seq library prep with PIPseq (particle-templated instant partition sequencing) chemistry coupled with the high sequencing capacity of the NovaSeq X Sequencing System can help improve scalability and sequencing efficiency.

Explore how single-cell RNA sequencing with PIPseq chemistry can reveal cell-level insights and discover new biomarkers with scalable, flexible single-cell sequencing workflows.



Learn how dual flow cell capability on the NovaSeq X Plus System can further improve efficiency and extend the discovery power of Perturb-Seq experiments.

A Perturb-Seq data analysis pipeline is defined by the following steps: 


1. Mapping and assignment: Raw sequencing reads are processed into usable, quantifiable data to identify the single-cell transcriptome, and the specific CRISPR gRNA that edited each individual cell. 


2. Quality control: Rigorous filtering is applied to remove multiplets, empty droplets, and cells where the specific CRISPR gRNA responsible for the edit is unclear. 


3. Phenotypic analysis: Differential gene expression and clustering analysis are performed on cells grouped by their genetic perturbation, ultimately providing insight into how each edit altered cellular function.7

Read depth requirements vary depending on several factors, including experimental design, transcriptome complexity, and perturbation efficiency. For Perturb-Seq experiments, recommended read depths range from tens of thousands to greater than 100,000 reads per cell, depending on requirements.

Yes, Perturb-Seq can be scaled for genome-wide screens. Advancements in direct–capture single-cell chemistry and high-throughput sequencing systems have helped alleviate historical throughput and cost bottlenecks that limited the use of Perturb-Seq in genome-scale studies.7

Read the Perturb-Seq for gene editing article to discover how researchers are using Perturb-Seq to obtain single-cell resolution of CRISPR-Cas9–perturbed cells more efficiently and at an unprecedented scale. 

Learn about the genome-wide sequencing capabilities of NovaSeq X Sequencing Systems. 


/ Results

Related resources

Multiomics

Solutions powered by NGS to enable your next multiomics profiling study and analysis.

Drug discovery and development

Innovative genomics solutions designed to inform drug discovery efforts and support development of targeted therapeutics.

Speak to a specialist

Talk to an expert to learn more about Perturb-Seq solutions.

References

  1. Sivakumar S, Wang Y, Goetsch SC, et al. Benchmarking and optimizing Perturb-seq in differentiating human pluripotent stem cells. Stem Cell Reports. 2025;20(12):102713. doi:10.1016/j.stemcr.2025.102713 
  2. Ota M, Spence JP, Zeng T, et al. Causal modelling of gene effects from regulators to programs to traits. Nature. 2026;650(8101):399-408. doi:10.1038/s41586-025-09866-3
  3. Rood JE, Hupalowska A, Regev A. Toward a foundation model of causal cell and tissue biology with a Perturbation Cell and Tissue Atlas. Cell. 2024;187(17):4520-4545. doi:10.1016/j.cell.2024.07.035 
  4. Shi H, Chi H. Next-generation CRISPR screens enable causal systems immunology. J Exp Med. 2026;223(3):e20241266. doi:10.1084/jem.20241266 
  5. Morris JA, Sun JS, Sanjana NE. Next-generation forward genetic screens: uniting high-throughput perturbations with single-cell analysis. Trends Genet. 2024;40(2):118-133. doi:10.1016/j.tig.2023.10.012 
  6. Yao D, Binan L, Bezney J, et al. Scalable genetic screening for regulatory circuits using compressed Perturb-seq. Nat Biotechnol. 2024;42(8):1282-1295. doi:10.1038/s41587-023-01964-9 
  7. Replogle JM, Saunders RA, Pogson AN, et al. Mapping information-rich genotype-phenotype landscapes with genome-scale Perturb-seq. Cell. 2022;185(14):2559-2575.e28. doi:10.1016/j.cell.2022.05.013