CRISPR Screening: Pooled vs Arrayed Screens and Libraries

Short answer: CRISPR screening is the use of a library of guide RNAs to perturb many genes in one experiment and then read out which perturbations change a phenotype. In a pooled screen, a mixed library is delivered to one population of cells and guide abundance is counted by sequencing; in an arrayed screen, each well holds a defined perturbation. Pooled screens are mainly used for discovery and arrayed screens mainly for validation and follow-up (Nature Reviews Methods Primers).

Definition and scope

A CRISPR screen replaces the single-locus edit of an ordinary genome-editing experiment with a library of guide RNAs. Addgene describes pooled screens as “far more complex than using CRISPR to modify a single genomic locus”. A library can target every gene in the genome or only a subset, such as one pathway. For the guide-level design questions, see sgRNA design; for the single-gene case, see gene knockout.

How a pooled screen runs

Addgene’s pooled-library guide describes the workflow in the following order. This is a description of the method as published, not a bench protocol.

1. Amplify and check the library. Addgene recommends next-generation sequencing (NGS) of the amplified library DNA to verify that it contains all of the expected plasmids.
2. Package as lentivirus (the guide describes 293T cells for packaging).
3. Transduce at a very low multiplicity of infection (MOI), so that each cell is likely to receive a single plasmid, and use many more cells than there are plasmids. Addgene’s example is a lab that used 200 times more cells than plasmids for the TKO library.
4. Apply the selection (see the next section).
5. Recover and count guides. Guide sequences are PCR-amplified from the cells and sequenced by NGS.

Positive versus negative selection

The two readouts collect different cells and compare different things.

Design	What happens to cells	What is collected	What is looked for
Positive screen	Most cells die under the selection	The survivors	Guides enriched among survivors. Addgene notes this can narrow tens of thousands of genes to several hundred or fewer.
Negative screen	Cells carrying essential-gene perturbations drop out	The selected arm and an untreated control	Guides depleted in the selected arm relative to the control

For analysis, the MAGeCK method (Li et al. 2014, Genome Biology) ranks sgRNAs, genes and pathways in genome-scale knockout screens and identifies positively and negatively selected genes at the same time. Its gene score uses a modified robust rank aggregation (alpha-RRA), which tests whether a gene’s sgRNAs rank consistently higher than expected under a uniform null. The authors illustrate it with examples such as EGFR in vemurafenib-treated BRAF-mutant A375 cells, and note that their evaluation used a limited number of public datasets and replicates.

Library types

Addgene groups its CRISPR libraries by the type of perturbation. Library sizes below are Addgene’s listed examples.

Type	Mechanism (per Addgene)	Listed example
Knockout	Indels at the target gene	Human Genome-Wide CRISPR Knockout (LGR): 20,367 genes, 4 gRNAs per gene, 81,776 gRNAs
Inhibition (CRISPRi)	dCas9 fused to a repressor	CRISPRi v2: 18,905 genes, 5 gRNAs per gene, 104,535 gRNAs
Activation (CRISPRa)	dCas9 fused to an activator	CRISPRa v2: 18,915 genes, 5 gRNAs per gene, 104,540 gRNAs
Base editing	Listed as its own library category	See base editing
Prime editing	Libraries of pegRNAs	Listed as its own category

Addgene also distinguishes one-plasmid libraries, where Cas is on the same plasmid, from two-plasmid libraries, which need cells or virus that already express Cas, and lists barcoded designs such as the iBAR library for high-MOI screening.

Pooled versus arrayed screens

Question	Pooled	Arrayed
Format	Whole library in one population; one tube of lentivirus in Thermo Fisher’s LentiPool description	One perturbation per well; Thermo Fisher’s LentiArray libraries use 96-well plates with up to four gRNAs per gene, each in its own well
Readout	Guide counts by NGS	Per-well assay
Typical role	Discovery (Nature Reviews Methods Primers)	Validation and follow-up (same source), which is more labour-intensive, costly and limited in scale
Instrumentation	Thermo Fisher states pooled use needs no high-throughput instrumentation	Designed for 96- and 384-well screening formats

Thermo Fisher notes that work can move from a pooled primary screen to an arrayed follow-up.

Pre-screen checklist drawn from the sources

• Has the amplified library been sequenced to confirm every expected guide is present?
• Is the MOI low enough that most infected cells carry one guide?
• Are there many more cells than library members?
• Is there both a selected arm and an untreated control for a negative screen?
• Is an analysis method such as MAGeCK chosen, and are replicates planned?
• Is an orthogonal follow-up planned for hits? For single-cell readouts of pooled perturbations, see Perturb-seq; for public dependency data to compare hits against, see DepMap.

Common misunderstandings

• A screen is not a list of validated genes. Pooled screens are discovery tools; hits are usually followed up in a smaller, often arrayed, format.
• “CRISPR screen” does not always mean knockout. Libraries exist for inhibition, activation, base editing and prime editing.
• More guides per gene is not the same as more cells. Library size sets how many cells must be carried through the screen.

FAQ

What is a CRISPR screen used for?

To test many genes at once for their effect on a selectable or measurable phenotype, such as survival under a drug.

What is the difference between pooled and arrayed CRISPR screens?

Pooled screens put the whole library into one cell population and count guides by sequencing; arrayed screens keep one perturbation per well. See the comparison table above.

How is the data analysed?

Guide counts are compared between conditions. MAGeCK is one published method for ranking sgRNAs and genes in knockout screens.

Sources

• Addgene, CRISPR Libraries, read 2026-11-08.
• Addgene, Pooled Libraries guide, read 2026-11-08.
• Li W et al., MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screens, Genome Biology 2014, PMC4290824, read 2026-11-08.
• Thermo Fisher Scientific, CRISPR libraries (manufacturer description), read 2026-11-08.
• CRISPR screens, Nature Reviews Methods Primers, doi:10.1038/s43586-021-00093-4.