CoChat
Sign Up
Cited by 465CRISPR and Genetic EngineeringSingle-cell and spatial transcriptomicsRNA Interference and Gene Delivery

Combinatorial single-cell CRISPR screens by direct guide RNA capture and targeted sequencing

Open access
Replogle, Joseph M., Norman, Thomas M., Xu, Albert, Hussmann, Jeffrey A., Chen, Jin, Cogan, J. Zachery, Meer, Elliott J., Terry, Jessica M. +9 more·2020·Nature Biotechnology
DOI 10.1038/s41587-020-0470-yPMC PMC7416462

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Single-cell CRISPR screens enable the exploration of mammalian gene function and genetic regulatory networks. However, use of this technology has been limited by reliance on indirect indexing of single-guide RNAs (sgRNAs). Here we present direct-capture Perturb-seq, a versatile screening approach in which expressed sgRNAs are sequenced alongside single-cell transcriptomes. Direct-capture Perturb-seq enables detection of multiple distinct sgRNA sequences from individual cells and thus allows pooled single-cell CRISPR screens to be easily paired with combinatorial perturbation libraries that contain dual-guide expression vectors. We demonstrate the utility of this approach for high-throughput investigations of genetic interactions and, leveraging this ability, dissect epistatic interactions between cholesterol biogenesis and DNA repair. Using direct capture Perturb-seq, we also show that targeting individual genes with multiple sgRNAs per cell improves efficacy of CRISPR interference and activation, facilitating the use of compact, highly active CRISPR libraries for single-cell screens. Last, we show that hybridization-based target enrichment permits sensitive, specific sequencing of informative transcripts from single-cell RNA-seq experiments.

Integrity signals

Registry

Crossref · 10.1038/s41587-020-0470-y

Cites

46 of 46 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 30 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Hybridization-based target enrichment permits sensitive, specific sequencing of informative transcripts from scRNA-seq experiments.

    “we show that hybridization-based target enrichment permits sensitive, specific sequencing of informative transcripts from single-cell RNA-seq experiments.”
    Quote in the paperSubstantial claim
  • Findingown data

    Capture of guides depended on sequence, as rates differed among guides according to their targeting sequence.

    “Unlike capture of GBCs, we found that guide capture was sequence-dependent with capture rates varying across guides by targeting sequence”
    Quote in the paperSubstantial claim
  • Findingown data

    Screening a dual-guide CRISPRi library in K562s by 3' direct capture Perturb-seq revealed adequate capture of guides from both vector positions (position A median 776 UMIs/cell; position B median 511 UMIs/cell).

    “Screening this library in K562s by 3’ direct capture Perturb-seq revealed adequate capture of guides from both vector positions (position A, median of 776 UMIs/cell; position B, median of 511 UMIs/cell; Supplementary Fig. 5b,c)”
    Quote in the paperSubstantial claim
  • Findingown data

    Random forest classifier prediction accuracies for classifying perturbed vs unperturbed cells were highly similar across platforms (r=0.91 for 3' sgRNA-CR1cs1; r=0.90 for 5' sgRNA-CR1).

    “prediction accuracies were highly similar across platforms (correlation with GBC Perturb-seq: r=0.91 for 3’ sgRNA-CR1cs1 capture, r=0.90 for 5’ sgRNA-CR1 capture)”
    Quote in the paperSubstantial claim
  • Findingnone given

    Direct-capture Perturb-seq sequences expressed sgRNAs together with single-cell transcriptomes.

    “Here we present direct-capture Perturb-seq, a versatile screening approach in which expressed sgRNAs are sequenced alongside single-cell transcriptomes.”
    Quote in the paperSubstantial claim
  • May combine two claims. 31
  • Classified as probably not a substantial claim. 10

  • May combine two claims.

    Findingown data

    Single-cell analysis showed that HUS1 plus FDPS perturbation created G2/M-arrested cells with much lower total mRNA, probably dying cells.

    “At the single-cell level, this generated a population of G2/M-arrested cells with notably decreased total mRNA content, likely representing dying cells”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The assignment method confidently identified guides in 84-94% of cells, versus 89% for GBC Perturb-seq, and guide distributions were approximately as expected on all platforms.

    “Nevertheless, our assignment procedure robustly assigned guide identities to 84–94% of cells (compared to 89% for GBC Perturb-seq) with roughly expected guide distributions across all platforms”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In iPSCs with Cas9, 3' direct capture Perturb-seq with pooled lentiviral packaging of 40 sgRNAs gave high guide capture rates (mean 999 UMIs/cell) and transcriptional phenotypes correlated for guides targeting the same gene.

    “In iPSCs, we again found high guide capture rates (mean capture of 999 UMIs/cell; Supplementary Fig. 4a,b) and transcriptional phenotypes that were correlated for guides targeting the same gene”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Genes acting at early and intermediate steps showed buffering: PMVK knockdown reduced the FDPS-specific response but left the shared cholesterol feedback response intact.

    “we observed that PMVK repression suppressed the FDPS-specific transcriptional response, while maintaining the cholesterol feedback response shared by both perturbations”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Screens using both direct capture platforms gave higher index capture than GBC-based method: 4.1-fold for 3' sgRNA-CR1cs1, 15.5-fold for 5' sgRNA-CR1cs1, 7.8-fold for 5' sgRNA-CR1; 3' capture of sgRNA-CR1cs2 had 0.56-fold.

    “At a constant sequencing depth, screens using both direct capture platforms gave higher index capture than the GBC-based method (4.1-fold higher for 3’ sgRNA-CR1cs1 capture; 15.5-fold higher for 5’ sgRNA-CR1cs1 capture; 7.8-fold higher for 5’ sgRNA-CR1 capture), with the exception of 3’ capture of sgRNA-CR1cs2, which had modestly lower capture (0.56-fold)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    HUS1 knockdown by itself caused only mild cell-cycle changes; with FDPS knockdown, cells largely bypassed the S-phase checkpoint and accumulated in G2/M.

    “While repression of HUS1 causes modest cell-cycle aberration alone, in combination with FDPS knockdown, we observed a substantial bypass of the S-phase checkpoint and an accumulation of cells in G2/M”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Gene expression modules conserved across platforms (cophenetic r=0.93 for 3' sgRNA-CR1cs1; r=0.95 for 5' sgRNA-CR1), with modules tending to cluster functionally based on regulation by three UPR branches.

    “we found gene expression modules that were conserved across platforms (cophenetic correlation with GBC Perturb-seq: r=0.93 for 3’ sgRNA-CR1cs1 capture; r=0.95 for 5’ sgRNA-CR1 capture), with modules tending to cluster functionally based on their regulation by the three UPR branches”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    After mapping guides to cells, >90% guide assignment was observed with >67% of guide-bearing cells expressing two sgRNAs.

    “we observed >90% guide assignment with >67% of guide-bearing cells expressing two sgRNAs, as expected given multiple infections, doublets from cell loading, and imperfect guide calling”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Clustering target genes by pseudo-bulk expression recovered known functional and physical interactions and matched GBC Perturb-seq relationships closely (cophenetic r=0.95 for 3' sgRNA-CR1cs1; r=0.95 for 5' sgRNA-CR1).

    “First, we hierarchically clustered target genes based on their pseudo-bulk expression profiles (Fig. 1e). This recapitulated known functional and physical interactions and, when compared to results generated with GBC Perturb-seq, produced highly similar relationships (cophenetic correlation with GBC Perturb-seq: r=0.95 for 3’ sgRNA-CR1cs1; r=0.95 for 5’ sgRNA-CR1).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Double perturbation of HUS1 and FDPS produced a neomorphic phenotype, with a transcriptional response absent after either single perturbation.

    “we found that these cells (with perturbation of both HUS1 and FDPS) demonstrate a neomorphic phenotype characterized by a transcriptional response not induced by either perturbation alone”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Strong target depletion was observed across platforms with median knockdown: 90% GBC, 94% 3' sgRNA-CR1cs1, 93% 3' sgRNA-CR1cs2, 95% 5' sgRNA-CR1, 93% 5' sgRNA-CR1cs1.

    “we found strong (and comparable) target depletion across platforms (median knockdown: 90% for GBC capture, 94% for 3’ sgRNA-CR1cs1 capture, 93% for 3’ sgRNA-CR1cs2 capture, 95% for 5’ sgRNA-CR1 capture, 93% for 5’ sgRNA-CR1cs1 capture)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Guide capture variation was consistent across screens and associated with the 5' nucleotides of guide RNAs, not with total GC content.

    “Importantly, this variation was correlated across screens and was related to the nucleotides at the 5’ ends of guide RNAs but not to overall GC content”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    At that sequencing depth, index library saturation was 0.75 for GBC, 0.96 for 3' sgRNA-CR1cs2, 0.71 for 3' sgRNA-CR1cs1, 0.28 for 5' sgRNA-CR1, and 0.60 for 5' sgRNA-CR1cs1.

    “At this sequencing depth, saturation of the index libraries is 0.75 for GBC, 0.96 for 3’ sgRNA-CR1cs2, 0.71 for 3’ sgRNA-CR1cs1, 0.28 for 5’ sgRNA-CR1, and 0.60 for 5’ sgRNA-CR1cs1.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Expression profiles after enrichment correlated strongly with unenriched profiles globally (r=0.98), per cell (median r=0.93), and per gene (median r=0.75).

    “Enriched gene expression profiles were highly correlated with unenriched profiles at the global (r=0.98), single-cell (median r=0.93), and single-gene (median r=0.75) levels”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The assignment rate in single-guide experiments was approximately 90%, on par with GBC Perturb-seq and CROP-seq.

    “given ~90% assignment rate in single-guide experiments (on par with GBC Perturb-seq and CROP-seq)”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Multiplexed sgRNAs nearly doubled CRISPR activity over the best single guide: CRISPRi median relative target expression 0.20 for sgRNAs 1+control vs 0.11 for sgRNAs 1+2 (Wilcoxon n=87, W=378, p=8e-11); CRISPRa median fold-activation 2.9 vs 4.7 (n=49, W=162, p=7e-6).

    “For both CRISPRi and CRISPRa, the multiplexed sgRNAs nearly doubled CRISPR activity over what was achieved with the best single guide (CRISPRi: sgRNAs 1+control, median relative target expression=0.20; sgRNAs 1+2, median relative target expression=0.11; Wilcoxon signed-rank two-sided test n=87 genes, W=378, p=8e-11; CRISPRa: sgRNAs 1+control, median fold-activation=2.9; sgRNAs 1+2, median fold-act”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    In the data, rates of novel sgRNA pairs varied from 0.09 to 0.15 across experiments, roughly consistent with a previous report.

    “In our data, we observed rates of novel pairs varying from 0.09–0.15 across experiments, which is roughly consistent with a previous report28.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [28] Horlbeck, M. A. et al. Mapping the genetic landscape of human cells. Cell 174, 953–967.e22 (2018). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    In the UPR direct capture Perturb-seq experiments, index alignment rates were 0.82 for GBCs, 0.35 for 3' sgRNA-CR1cs2, 0.62 for 3' sgRNA-CR1cs1, 0.71 for 5' sgRNA-CR1, and 0.62 for 5' sgRNA-CR1cs1.

    “We observed index alignment rates of 0.82 for GBCs, 0.35 for 3’ sgRNA-CR1cs2, 0.62 for 3’ sgRNA-CR1cs1, 0.71 for 5’ sgRNA-CR1, and 0.62 for 5’ sgRNA-CR1cs1 in our UPR direct capture Perturb-seq experiments.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    Direct capture Perturb-seq can identify cells bearing undesired sgRNA pairs generated from intermolecular lentiviral recombination and computationally exclude them from downstream analysis.

    “Lastly, with direct capture Perturb-seq, we can identify cells bearing undesired sgRNA pairs (generated from intermolecular lentiviral recombination between programmed pairs9–12) and computationally exclude them from downstream analysis.”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [9] Hill, A. J. et al. On the design of CRISPR-based single-cell molecular screens. Nat. Methods 15, 271 (2018). source

      Not checked in the free run.

    • cites [10] Adamson, B., Norman, T. M., Jost, M. & Weissman, J. S. Approaches to maximize sgRNA-barcode coupling in Perturb-seq screens. Preprint at bioRxiv https://doi.org/10.1101/298349 (2018). source

      Not checked in the free run.

    • cites [11] Xie, S., Cooley, A., Armendariz, D., Zhou, P. & Hon, G. C. Frequent sgRNA-barcode recombination in single-cell perturbation assays. PLoS ONE 13, e0198635 (2018). source

      Not checked in the free run.

    • cites [12] Feldman, D., Singh, A., Garrity, A. J. & Blainey, P. C. Lentiviral co-packaging mitigates the effects of intermolecular recombination and multiple integrations in pooled genetic screens. Preprint at bioRxiv https://doi.org/10.1101/262121 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    The mixture-model guide assignment procedure produced a coherent proportion of cells assigned to each guide identity and a coherent multiplet rate across platforms, within 1-1.6 fold of expectations based on library transduction and published multiple encapsulation rates.

    “This procedure produced a coherent proportion of cells assigned to each guide identity (Supplementary Fig. 2h) and a coherent multiplet rate across platforms–within 1–1.6 fold of expectations based on library transduction (assuming Poisson infection distribution) and published multiple encapsulation rates (Fig. 1d).”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Infection follows a Poisson distribution.

  • May combine two claims.

    Study design and conductreasoning

    Direct capture allows unbiased assignment of sgRNAs to cells, enabling identification of novel sgRNA pairs, which were excluded from downstream analysis.

    “because direct capture allows us to assign sgRNAs to cells in an unbiased way, we were able to identify novel sgRNA pairs and excluded them from downstream analysis.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The capture sequences do not distinguish guides and could allow simultaneous capture of other features, including antibodies and oligo-tagged markers.

    “However, these capture sequences are not guide-specific, and thus, in principle, will enable multiplexed capture of additional features, such as antibodies22,23 and other oligo-tagged markers24.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [22] Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in single cells. Nat. Methods 14, 865–868 (2017). source

      Not checked in the free run.

    • cites [23] Peterson, V. M. et al. Multiplexed quantification of proteins and transcripts in single cells. Nat. Biotechnol. 35, 936–939 (2017). source

      Not checked in the free run.

    • cites [24] Zhang, S.-Q. et al. High-throughput determination of the antigen specificities of T cell receptors in single cells. Nat. Biotechnol. 36, 1156–1159 (2018). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    An earlier model held that repressing cholesterol biosynthesis causes toxic intermediates to accumulate, producing replication stress and arrest after genotoxin exposure.

    “We previously proposed a model for GIs between cholesterol biosynthesis and DNA repair genes wherein repression of the former leads to the buildup of toxic metabolic intermediates in cells, which then cause replicative stress and genotoxin-activated cell cycle arrest28.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [28] Horlbeck, M. A. et al. Mapping the genetic landscape of human cells. Cell 174, 953–967.e22 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    CROP-seq delivers pooled guide libraries while keeping sgRNAs paired with their polyadenylated indexes.

    “In this way, CROP-seq ensures delivery of pooled guide libraries to cells with faithful pairing of sgRNAs and polyadenylated “indexes”.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [8] Datlinger, P. et al. Pooled CRISPR screening with single-cell transcriptome readout. Nat. Methods 14, 297–301 (2017). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Current single-cell CRISPR screening implementations have limitations including poor scalability, dependence on specialized vector systems, and high cost.

    “current implementations face technical and practical limitations that unnecessarily restrict their use. Here, we present advances that address these limitations, specifically poor scalability, dependence on specialized vector systems and high cost9–12”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [9] Hill, A. J. et al. On the design of CRISPR-based single-cell molecular screens. Nat. Methods 15, 271 (2018). source

      Not checked in the free run.

    • cites [10] Adamson, B., Norman, T. M., Jost, M. & Weissman, J. S. Approaches to maximize sgRNA-barcode coupling in Perturb-seq screens. Preprint at bioRxiv https://doi.org/10.1101/298349 (2018). source

      Not checked in the free run.

    • cites [11] Xie, S., Cooley, A., Armendariz, D., Zhou, P. & Hon, G. C. Frequent sgRNA-barcode recombination in single-cell perturbation assays. PLoS ONE 13, e0198635 (2018). source

      Not checked in the free run.

    • cites [12] Feldman, D., Singh, A., Garrity, A. J. & Blainey, P. C. Lentiviral co-packaging mitigates the effects of intermolecular recombination and multiple integrations in pooled genetic screens. Preprint at bioRxiv https://doi.org/10.1101/262121 (2018). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Pooling ratios were designed to ensure even representation among targeting guides and to deliver the two non-targeting controls at 4-fold excess.

    “Pooling ratios were designed to ensure even representation among targeting guides and delivery of sgNegCtrl2 and sgNegCtrl3 at 4-fold excess.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Guides selected for the multiplexed libraries were predicted to be highly active (top two by rank in hCRISPRi v2.1 and hCRISPRa v2).

    “We selected guide RNA targeting sequences predicted to be highly active (the top two by rank in hCRISPRi v2.1 and hCRISPRa v230)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [30] Moreno, A. M. et al. In situ gene therapy via AAV-CRISPR-Cas9-mediated targeted gene regulation. Mol. Ther. 26, 1818–1827 (2018). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Cell viability for all three UPR libraries remained above 87% up to the point of sorting.

    “Up to this point, cell viability for all three libraries remained >87%.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The multiplexed guides exceeded predictions from a dominant model, indicating possible synergy between them.

    “Moreover, in both cases, the multiplexed sgRNAs appeared to perform better than expected based on a dominant model of guide activity, suggesting some degree of synergy between multiplexed sgRNAs”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The loading scheme increased doublets but minimized reagent cost per recovered cell.

    “Our loading scheme (designed to recover ~15,000 cells per lane) increased these doublets but notably also minimized reagent cost per recovered cell.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationnone given

    Cells with more than two guides may arise from multiple infection events or double loading into droplets.

    “Cells with more than two guides, on the other hand, may arise from either multiple infection events or double loading into droplets.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Libraries that are compact and highly active because they express several sgRNAs per gene can scale direct-capture Perturb-seq, covering more genes and reducing false negatives from weak expression changes.

    “These results show that compact, highly active libraries (expressing multiple sgRNAs per gene) can be used to scale single-cell experiments with direct capture Perturb-seq, interrogating more genes while minimizing false negatives due to insufficient expression modulation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    The 5' direct-capture platform works with standard sgRNAs, captures guides more efficiently, and supports V(D)J clonotype analysis, while the 3' platform suits molecular recording and lineage tracing.

    “For example, 5’ direct capture is compatible with standard sgRNAs, has higher guide capture rates, and allows for V(D)J clonotype analysis, whereas 3’ direct capture is compatible with many molecular recording and lineage tracing approaches37.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [37] Chan, M. M. et al. Molecular recording of mammalian embryogenesis. Nature 570, 77–82 (2019). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    For GBC Perturb-seq, a threshold separating the bimodal distribution of GBC coverage (reads per UMI) was used as previously described.

    “For GBC Perturb-seq, we did this using a threshold that separates the bimodal distribution of GBC coverage (reads per UMI) as previously described4.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [4] Adamson, B. et al. A multiplexed single-cell CRISPR screening platform enables systematic dissection of the unfolded protein response. Cell. 167, 1867–1882.e21 (2016). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Since enrichment acts on finished libraries, target genes need not be chosen in advance and can be adjusted repeatedly within one experiment.

    “because our target enrichment procedure is performed on final libraries, target genes do not need to be selected a priori and can be iteratively refined for a single experiment.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Cell pools for the UPR experiments were more than 94% BFP+ at the time of droplet encapsulation.

    “cell pools >94% BFP+”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The three UPR CRISPRi libraries each contain 32 guide RNA targeting regions: 30 target genes previously reported to activate the unfolded protein response upon depletion, and 2 are non-targeting controls.

    “Each of these libraries encodes guide RNAs programmed with 32 unique guide RNA targeting regions: 30 which target genes whose depletion was previously shown to activate the unfolded protein response (UPR) by GBC Perturb-seq4 and 2 non-targeting controls (sgNegCtrl2 and sgNegCtrl3).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [4] Adamson, B. et al. A multiplexed single-cell CRISPR screening platform enables systematic dissection of the unfolded protein response. Cell. 167, 1867–1882.e21 (2016). source

      Not checked in the free run.

37 of 37

  • The paper relies on this without support.

    CROP-seq cannot deliver multiple sgRNAs because of limits on cassette size.

    “However, due to constraints on cassette size, CROP-seq is thought to be incompatible with delivery of multiple sgRNAs.”

  • The paper relies on this without support.

    GBC Perturb-seq cannot be scaled.

    “However, GBC Perturb-seq is not scalable.”

  • The paper relies on this without support.

    Current droplet-based scRNA-seq methods require whole-transcriptome sequencing for phenotyping, which can be too expensive.

    “Lastly, we addressed the fact that current droplet-based scRNA-seq implementations are constrained by sequencing the whole transcriptome for phenotyping, which can be prohibitively expensive.”

  • The authors state this as an assumption.

    Guides are independently paired.

    “as it assumes that guides are independently paired”

  • The authors state this as an assumption.

    Infection follows a Poisson distribution.

    “assuming Poisson infection distribution”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • The 3′ sgRNA-CR1cs1 experiment is labeled as GBC Perturb-seq in Methods

    Minor

    The Methods call the sgRNA-CR1cs1 library's 3′ lane a GBC experiment, whereas the results identify its measurements as 3′ direct guide capture.

1 of 1

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Replogle, J. M., Norman, T. M., Xu, A., Hussmann, J. A., Chen, J., Cogan, J. Z., Meer, E. J., Terry, J. M., Riordan, D. P., Srinivas, N., Fiddes, I. T., Arthur, J. G., Alvarado, L. J., Pfeiffer, K. A., Mikkelsen, T. S., Weissman, J. S., & Adamson, B.
Affiliations
  • Howard Hughes Medical Institute (US)
  • Memorial Sloan Kettering Cancer Center (US)
  • 10X Genomics (United States) (US)
  • Princeton University (US)
Corresponding author
  • Weissman, Jonathan S.
  • Adamson, Britt
ORCID
Venue
Nature Biotechnology
Publisher
Springer Science and Business Media LLC
Published
2020-03-30
Volume
38
Issue
8
Pages
954-961
ISSN
  • 0733-222X (ISSN-L)
  • 1087-0156
  • 1546-1696
DOI
10.1038/s41587-020-0470-y
PMC
PMC7416462
OpenAlex
W3013900499
Semantic Scholar
fde2e40ef51ca9eab137402cb38855d3092fabe2
Cited by
465
References
46
Citations by year
2026: 89 · 2025: 100 · 2024: 79 · 2023: 61 · 2022: 60 · 2021: 49 · 2020: 23 · 2019: 4
FWCI
15.74
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Biotechnology (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nd)
  • PubMed (publishedVersion)
  • PubMed Central (acceptedVersion)
  • eScholarship (California Digital Library) (submittedVersion)
Funders
  • U.S. Department of Health & Human Services | National Institutes of Health (F31 NS115380, GM134154, P50 GM102706, U01 CA168370, R01 DA036858, RM1HG009490)
  • Damon Runyon Cancer Research Foundation (DRG-[2211-15], DRG-2262-16)
  • Howard Hughes Medical Institute
  • Jane Coffin Childs Memorial Fund for Medical Research
  • 10x Genomics
  • United States Department of Defense | Defense Advanced Research Projects Agency (HR0011-19-2-0007)
  • Chan Zuckerberg Initiative
  • Princeton University
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › CRISPR and Genetic Engineering
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA Interference and Gene Delivery
Keywords
single-cell CRISPR screens, Perturb-seq, epistasis, DNA repair, CRISPR interference, CRISPR activation, single-cell transcriptomics
MeSH terms
RNA, Guide, CRISPR-Cas Systems, Gene Expression Regulation, Humans, RNA, Guide, Kinetoplastida, Gene Targeting, Nucleic Acid Amplification Techniques, HEK293 Cells, Single-Cell Analysis, High-Throughput Nucleotide Sequencing, Transcriptome, CRISPR-Cas Systems

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-04.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-04.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review