CoChat
Sign Up
Cited by 1,051Single-cell and spatial transcriptomicsBioinformatics and Genomic NetworksGenomics and Chromatin Dynamics

Best practices for single-cell analysis across modalities

Open access
Heumos, Lukas, Schaar, Anna C., Lance, Christopher, Litinetskaya, Anastasia, Drost, Felix, Zappia, Luke, Lücken, Malte D., Strobl, Daniel C., Henao, Juan, Curion, Fabiola, Single-cell Best Practices Consortium, Aliee, Hananeh, Ansari, Meshal, Badia-i-Mompel, Pau, Büttner, Maren, Dann, Emma, Dimitrov, Daniel, Dony, Leander, Frishberg, Amit, He, Dongze, Hediyeh-zadeh, Soroor, Hetzel, Leon, Ibarra, Ignacio L., Jones, Matthew G., Lotfollahi, Mohammad, Martens, Laura D., Müller, Christian L., Nitzan, Mor, Ostner, Johannes, Palla, Giovanni, Patro, Rob, Piran, Zoe, Ramírez-Suástegui, Ciro, Saez-Rodriguez, Julio, Sarkar, Hirak, Schubert, Benjamin, Sikkema, Lisa, Srivastava, Avi, Tanevski, Jovan, Virshup, Isaac, Weiler, Philipp, Schiller, Herbert B., Theis, Fabian J.·2023·Nature Reviews Genetics
DOI 10.1038/s41576-023-00586-wPMC PMC10066026
Open this paper in CoChat — free accountOpen PDF

Abstract

Recent advances in single-cell technologies have enabled high-throughput molecular profiling of cells across modalities and locations. Single-cell transcriptomics data can now be complemented by chromatin accessibility, surface protein expression, adaptive immune receptor repertoire profiling and spatial information. The increasing availability of single-cell data across modalities has motivated the development of novel computational methods to help analysts derive biological insights. As the field grows, it becomes increasingly difficult to navigate the vast landscape of tools and analysis steps. Here, we summarize independent benchmarking studies of unimodal and multimodal single-cell analysis across modalities to suggest comprehensive best-practice workflows for the most common analysis steps. Where independent benchmarks are not available, we review and contrast popular methods. Our article serves as an entry point for novices in the field of single-cell (multi-)omic analysis and guides advanced users to the most recent best practices.

Integrity signals

Registry

Crossref · 10.1038/s41576-023-00586-w

Cites

263 of 263 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 146 cited sources checked · 0 supported · 0 flagged

  • Premisecitation

    Removing ambient RNA as an initial quality-control step improves downstream analyses for many tissues.

    “one should consider removing ambient RNA as an initial analysis step in quality control to improve downstream analyses for many tissues21–23.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [21] Young, M. D. & Behjati, S. SoupX removes ambient RNA contamination from droplet-based single-cell RNA sequencing data. Gigascience 9, giaa151 (2020). source

      Not checked in the free run.

    • cites [22] Fleming, S. J. et al. Unsupervised removal of systematic background noise from droplet-based single-cell experiments using CellBender. bioRxiv https://doi.org/10.1101/791699 (2022). source

      Not checked in the free run.

    • cites [23] Yang, S. et al. Decontamination of ambient RNA in single-cell RNA-seq with DecontX. Genome Biol. 21, 57 (2020). source

      Not checked in the free run.

  • Premisecitation

    Using counts per million as input with the shifted logarithm transformation reflects unrealistically large overdispersion.

    “The shifted logarithm transformation with size factor s performs well but should not be used with counts per million as an input, as it reflects an unrealistically large overdispersion.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [29] Ahlmann-Eltze, C. & Huber, W. Comparison of transformations for single-cell RNA-seq data. bioRxiv https://doi.org/10.1101/2021.06.24.449781 (2022). source

      Not checked in the free run.

  • Premisecitation

    Existing computational best practices and tool recommendations for scRNA-seq are outdated or incomplete.

    “although computational best practices and tool recommendations have previously been outlined for scRNA-seq2,5,6,17, they are either outdated or incomplete.”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [2] Amezquita, R. A. et al. Orchestrating single-cell analysis with bioconductor. Nat. Methods 17, 137–145 (2020). source

      Not checked in the free run.

    • cites [5] Luecken, M. D. & Theis, F. J. Current best practices in single-cell RNA-seq analysis: a tutorial. Mol. Syst. Biol. 15, e8746 (2019). This scRNA-seq-focused review and workflow template served as inspiration for the current benchmark-based Expert Recommendation. source

      Not checked in the free run.

    • cites [6] Kharchenko, P. V. The triumphs and limitations of computational methods for scRNA-seq. Nat. Methods 18, 723–732 (2021). The article highlights limitations of computational methods, which analysts should be aware of. source

      Not checked in the free run.

    • cites [17] Stuart, T. & Satija, R. Integrative single-cell analysis. Nat. Rev. Genet. 20, 257–272 (2019). A review of the advent of multimodal single-cell data with a focus on the experimental assays and data integration. source

      Not checked in the free run.

  • Premisecitation

    Ambient RNA contamination can cause cell-type-specific marker gene transcripts to be detectable in other cell populations, blending populations together.

    “Ambient RNA contamination can lead to cell-type-specific marker gene transcripts being detectable also in other cell populations, which can blend different cell populations together21.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [21] Young, M. D. & Behjati, S. SoupX removes ambient RNA contamination from droplet-based single-cell RNA sequencing data. Gigascience 9, giaa151 (2020). source

      Not checked in the free run.

  • Premisecitation

    The shifted logarithm transformation with size factor s performs well in a benchmark of 22 transformations.

    “The shifted logarithm transformation with size factor s performs well but should not be used with counts per million as an input, as it reflects an unrealistically large overdispersion.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [29] Ahlmann-Eltze, C. & Huber, W. Comparison of transformations for single-cell RNA-seq data. bioRxiv https://doi.org/10.1101/2021.06.24.449781 (2022). source

      Not checked in the free run.

  • Classified as probably not a substantial claim. 24
  • May combine two claims. 45

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

    Premisecitation

    More than 1,400 computational tools are currently available for analysing scRNA-seq data.

    “leading to more than 1,400 tools currently being available to analyse scRNA-seq data1”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Zappia, L. & Theis, F. J. Over 1000 tools reveal trends in the single-cell RNA-seq analysis landscape. Genome Biol. 22, 301 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    scDblFinder outperforms other doublet detection methods in accuracy and computational efficiency.

    “Several benchmarks have highlighted that scDblFinder outperforms other methods in terms of doublet detection accuracy and computational efficiency25–27.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [25] Xi, N. M. & Li, J. J. Protocol for executing and benchmarking eight computational doublet-detection methods in single-cell RNA sequencing data analysis. Star. Protoc. 2, 100699 (2021). source

      Not checked in the free run.

    • cites [26] Xi, N. M. & Li, J. J. Benchmarking computational doublet-detection methods for single-cell RNA sequencing data. Cell Syst. 12, 176–194.e6 (2021). source

      Not checked in the free run.

    • cites [27] Neavin, D. et al. Demuxafy: improvement in droplet assignment by integrating multiple single-cell demultiplexing and doublet detection methods. bioRxiv https://doi.org/10.1101/2022.03.07.483367 (2022). source

      Not checked in the free run.

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

    Premisecitation

    Applying multiple doublet detection methods and comparing their results can increase doublet detection accuracy.

    “Additionally, it can be beneficial to apply multiple doublet detection methods and compare the results to increase the accuracy of doublet detection27.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [27] Neavin, D. et al. Demuxafy: improvement in droplet assignment by integrating multiple single-cell demultiplexing and doublet detection methods. bioRxiv https://doi.org/10.1101/2022.03.07.483367 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Leiden yields guaranteed connected communities and is computationally more efficient than Louvain.

    “Louvain’s successor Leiden circumvents this issue by yielding guaranteed connected communities and is computationally more efficient52.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [52] Traag, V. A., Waltman, L. & van Eck, N. J. From Louvain to Leiden: guaranteeing well-connected communities. Sci. Rep. 9, 5233 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    RNA velocity cannot cover all time scales because it models splicing kinetics, so the observed process must occur during that time frame.

    “As it is splicing kinetics that are modelled, the observed process must also occur during this time frame70.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [70] Bergen, V., Soldatov, R. A., Kharchenko, P. V. & Theis, F. J. RNA velocity-current challenges and future perspectives. Mol. Syst. Biol. 17, e10282 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Slingshot performed better for simple topologies, whereas PAGA and RaceID/StemID scored better for complex trajectories.

    “Although Slingshot62 performed better for simple topologies, PAGA63 and RaceID/StemID64 scored better for complex trajectories65.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [65] Saelens, W., Cannoodt, R., Todorov, H. & Saeys, Y. A comparison of single-cell trajectory inference methods: towards more accurate and robust tools. Nat. Biotechnol. 37, 547–554 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    scANVI can incorporate cell-type labels, which helps conserve biological variation.

    “scANVI40 can incorporate the cell-type labels, which is favourable as it can help to conserve biological variation35.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [35] Luecken, M. D. et al. Benchmarking atlas-level data integration in single-cell genomics. Nat. Methods 19, 41–50 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Analytical Pearson residuals are better suited for selecting biologically variable genes and identifying rare cell identities.

    “analytical Pearson residuals are better suited for selection of biologically variable genes and identification of rare cell identities31.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [31] Lause, J., Berens, P. & Kobak, D. Analytic Pearson residuals for normalization of single-cell RNA-seq UMI data. Genome Biol. 22, 258 (2021). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    If a gene includes multiple pronounced kinetics, lineage-specific models are more appropriate.

    “Moreover, if a gene includes multiple, pronounced kinetics, lineage-specific models are more appropriate69.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [69] Weiler, P., Van den Berge, K., Street, K. & Tiberi, S. A guide to trajectory inference and RNA velocity. Methods Mol. Biol. 2584, 269–292 (2023). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Deviance performed favourably for identifying genes with high variance across subpopulations and for selecting informative genes.

    “It performed favourably for identifying genes with high variance across subpopulations and thus for selecting informative genes, as shown in an independent comparison20.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [20] Germain, P.-L., Sonrel, A. & Robinson, M. D. pipeComp, a general framework for the evaluation of computational pipelines, reveals performant single cell RNA-seq preprocessing tools. Genome Biol. 21, 227 (2020). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisereasoning

    Under the assumption of constant rates, phase portraits form an almond shape with induction and repression phases.

    “Under the assumption of constant rates, phase portraits form an almond shape with induction (upper half/arc) and repression (lower half/arc) phases.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: RNA velocity inference assumes gene independence and constant rates of transcription, splicing and degradation.

  • May combine two claims.

    Premisereasoning

    Ranking genes by deviance is performed on raw counts and is therefore not sensitive to normalization.

    “Additionally, ranking genes by deviance is performed on raw counts and is therefore not sensitive to normalization.”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    Gene expression is controlled by a complex interplay of regulatory mechanisms, including epigenetics and chromatin accessibility.

    “Gene expression is controlled by a complex interplay of regulatory mechanisms, including epigenetics and chromatin accessibility132.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [132] Isbel, L., Grand, R. S. & Schübeler, D. Generating specificity in genome regulation through transcription factor sensitivity to chromatin. Nat. Rev. Genet. 23, 728–740 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Single-cell-specific differential gene expression methods systematically underestimate gene expression variance and are prone to wrongly labelling highly expressed genes as differentially expressed.

    “Single-cell-specific methods were found to systematically underestimate the variance of gene expression and to be prone to wrongly labelling highly expressed genes as differentially expressed86.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [86] Squair, J. W. et al. Confronting false discoveries in single-cell differential expression. Nat. Commun. 12, 5692 (2021). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Perturbation modelling tools have not yet been independently benchmarked.

    “for which tools have not yet been independently benchmarked.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    LIANA provides an overall ranking for several combinations of method and database.

    “Owing to the lack of consensus between tools, we recommend using LIANA, which provides an overall ranking for several combinations of method and database126.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [126] Dimitrov, D. et al. Comparison of methods and resources for cell-cell communication inference from single-cell RNA-seq data. Nat. Commun. 13, 3224 (2022). source

      Not checked in the free run.

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

    Premisecitation

    scATAC-seq measures genome-wide chromatin accessibility in individual cells.

    “single-cell assay for transposase-accessible chromatin sequencing (scATAC-seq) measures genome-wide chromatin accessibility in individual cells133,134”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [133] Cusanovich, D. A. et al. Multiplex single-cell profiling of chromatin accessibility by combinatorial cellular indexing. Science 348, 910–914 (2015). source

      Not checked in the free run.

    • cites [134] Buenrostro, J. D. et al. Single-cell chromatin accessibility reveals principles of regulatory variation. Nature 523, 486–490 (2015). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Reducing K for the KNN graph or constructing a graph on particular lineages mitigates the loss of information in KNN-based methods.

    “Reducing K for the KNN graph or constructing a graph on particular lineages mitigates this issue112.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [112] Dann, E., Henderson, N. C., Teichmann, S. A., Morgan, M. D. & Marioni, J. C. Differential abundance testing on single-cell data using k-nearest neighbor graphs. Nat. Biotechnol. 40, 245–253 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Most scATAC-seq workflows use a cell-by-peak or cell-by-bin matrix, which performs better than matrices of gene or TF motif features.

    “Most workflows use a cell-by-peak or cell-by-bin matrix as a basis for analysis, which performs better than matrices of gene or TF motif features135”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [135] Chen, H. et al. Assessment of computational methods for the analysis of single-cell ATAC-seq data. Genome Biol. 20, 241 (2019). source

      Not checked in the free run.

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

    Premisecitation

    Existing benchmarks on bulk ATAC-seq data recommend edgeR for DAR determination when sample size is limited and DESeq2 for large sample sizes.

    “Although the performance on scATAC-seq data has not been benchmarked yet, existing benchmarks on bulk ATAC-seq data recommend edgeR for the determination of DARs when sample size is limited and DESeq2 in the case of large sample sizes146.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [146] Gontarz, P. et al. Comparison of differential accessibility analysis strategies for ATAC-seq data. Sci. Rep. 10, 10150 (2020). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    SCENIC can infer active TFs, and matching corresponding motifs with peak regions adds interpretability.

    “This approach can be extended by inferring active TFs using SCENIC followed by matching the corresponding motifs with peak regions to add additional interpretability156.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [156] Liu, L. et al. Deconvolution of single-cell multi-omics layers reveals regulatory heterogeneity. Nat. Commun. 10, 470 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The cell-by-peak matrix is more sparse than scRNA-seq data, with only 1-10% of peaks called per cell because diploid cells have only two copies of assayable chromatin.

    “Notably, the cell-by-peak matrix is even more sparse than scRNA-seq data, with only 1–10% of peaks called in each cell owing to the presence of only two copies of assayable chromatin in cells of a diploid organism135.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [135] Chen, H. et al. Assessment of computational methods for the analysis of single-cell ATAC-seq data. Genome Biol. 20, 241 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    TF-motif enrichment facilitates characterization of cell identity and can be conducted at cluster level using a hypergeometric test on cluster-specific DARs.

    “TF-motif enrichment facilitates the characterization of cell identity and can be conducted on a cluster level using a hypergeometric test on cluster-specific DARs140.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [140] Granja, J. M. et al. ArchR is a scalable software package for integrative single-cell chromatin accessibility analysis. Nat. Genet. 53, 403–411 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Identifying peaks requires a sufficient number of cells and may fail in rare cell types.

    “Identifying peaks requires a sufficient number of cells and therefore may fail in rare cell types136.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [136] Fang, R. et al. Comprehensive analysis of single cell ATAC-seq data with SnapATAC. Nat. Commun. 12, 1337 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Although paired measurements of many modality combinations are possible, different modalities are still commonly measured independently, resulting in unpaired data.

    “Although advancements in experimental assays allow for the paired measurements of many modality combinations246, different modalities are still commonly measured independently, resulting in unpaired data247.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [246] Mimitou, E. P. et al. Scalable, multimodal profiling of chromatin accessibility, gene expression and protein levels in single cells. Nat. Biotechnol. 39, 1246–1258 (2021). A review on multimodal data integration that further categorizes methods into different types of data integration task as for example used here. source

      Not checked in the free run.

    • cites [247] Lake, B. B. et al. Integrative single-cell analysis of transcriptional and epigenetic states in the human adult brain. Nat. Biotechnol. 36, 70–80 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Binarization may remove biological information, so direct modelling of scATAC counts has been suggested.

    “However, this may also remove biological information and therefore modelling of scATAC counts directly has been suggested142.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [142] Martens, L. D., Fischer, D. S., Theis, F. J. & Gagneur, J. Modeling fragment counts improves single-cell ATAC-seq analysis. bioRxiv https://doi.org/10.1101/2022.05.04.490536 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Calling peaks within clusters improves peak detection sensitivity and reduces the risk of missing peaks in rare cell types masked by noise from abundant cell types.

    “The sensitivity of peak detection can be improved by calling them within clusters, which reduces the risk of missing peaks in rare cell types masked by the noise of other highly abundant cell types.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [136] Fang, R. et al. Comprehensive analysis of single cell ATAC-seq data with SnapATAC. Nat. Commun. 12, 1337 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Simple models are often sufficient for cell-type annotation, and smoothing gene activity scores among neighbouring cells, often using MAGIC, can enhance visualization.

    “To guide cell-type annotation, simple models are often sufficient, and visualization can be enhanced by smoothing gene activity scores among neighbouring cells, which is often performed using MAGIC152.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [152] van Dijk, D. et al. Recovering gene interactions from single-cell data using data diffusion. Cell 174, 716–729.e27 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Low-quality scATAC-seq cells often form a cluster with low counts and low TSS enrichment scores and should be removed.

    “Low-quality cells often form a cluster combining low counts and low TSS enrichment scores that should be removed137.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [137] Ou, J. et al. ATACseqQC: a Bioconductor package for post-alignment quality assessment of ATAC-seq data. BMC Genomics 19, 169 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisenone given

    totalVI and MultiVI can be used for mosaic integration but are applicable only to CITE-seq and Multiome data, respectively.

    “Although totalVI and MultiVI can also be used for mosaic integration, they are both applicable only to CITE-seq and Multiome data, respectively.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    GLUE works well for integrating more than two modalities and won the NeurIPS 2021 multimodal single-cell data integration challenge.

    “It has been shown to work well for the integration of more than two modalities and is the winner of the NeurIPS 2021 multimodal single-cell data integration challenge252.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [252] Lance, C. et al. Multimodal single cell data integration challenge: results and lessons learned. in Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track (eds Kiela, D., Ciccone, M. & Caputo, B.) vol. 176 162–176 (PMLR, 2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    DSB uses background droplets representing protein background noise to correct values in cells while removing cell-to-cell variation by combining isotype control levels with the specific background level of each cell.

    “DSB uses background droplets that represent protein background noise to correct values in cells while removing cell-to-cell variation by combining isotype control levels with the specific background level of the respective cell.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [161] Mulè, M. P., Martins, A. J. & Tsang, J. S. Normalizing and denoising protein expression data from droplet-based single cell profiling. Nat. Commun. 13, 2099 (2022). source

      Not checked in the free run.

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

    Premisereasoning

    Somatic V(D)J recombination generates highly diverse AIR sequences across the population of B and T cells in an individual to capture a vast range of antigens.

    “To capture the vast range of antigens, somatic V(D)J recombination generates highly diverse AIR sequences across the population of B and T cells in an individual”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Cells with more than two assignments for either VJ or VDJ chains are always indicative of doublets.

    “However, cells with more than two assignments for either VJ or VDJ chains are always indicative of doublets.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Premisenone given

    Private clonotypes represent patient-specific clonal responses that might be valuable for personalized medicine.

    “By contrast, private clonotypes represent patient-specific clonal responses that might be valuable for personalized medicine.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisenone given

    Clonotypes can be identified by identical V gene and identical nucleic acid sequences for VJ and VDJ CDR3 for TCRs, or based on distance for BCR lineage reconstruction accounting for somatic hypermutation.

    “Clonotypes can be identified by identical V gene and identical nucleic acid sequences for VJ and VDJ CDR3 for TCRs or based on distance as implemented in the analysis frameworks for lineage reconstruction of BCRs accounting for somatic hypermutation”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Premisenone given

    Public clonotypes appear in more than one donor and can represent shared immunological response.

    “Public clonotypes appear in more than one donor and can represent shared immunological response.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Lymphocytes can express dual AIRs, with about 10% expressing multiple VJ chains paired with a single VDJ chain.

    “Lymphocytes can express dual AIRs173 with ~10% expressing multiple VJ chains paired with a single VDJ chain.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [173] Schuldt, N. J. & Binstadt, B. A. Dual TCR T cells: identity crisis or multitaskers? J. Immunol. 202, 637–644 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Variable antibody efficacy can cause strong batch effects when integrating ADT data across studies, and these should be corrected.

    “As antibody efficacy is variable, the integration of ADT data across several studies can lead to strong batch effects that should be corrected for160.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [160] Zheng, Y., Jun, S.-H., Tian, Y., Florian, M. & Gottardo, R. Robust normalization and integration of single-cell protein expression across CITE-seq datasets. Preprint at https://doi.org/10.1101/2022.04.29.489989. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Recombination of V(D)J gene segments and imprecise junction of V and J segments produce the CDR3 region in VJ and VDJ chains that is mainly responsible for AIR-antigen binding.

    “The recombination of V(D)J gene segments and the imprecise junction of V and J segments produce the CDR3 region in VJ and VDJ chains that is mainly responsible for AIR–antigen binding.”
    Quote in the paperNot a substantial claim

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

    Premisenone given

    BCRs directly recognize soluble or membrane-bound epitopes, whereas TCRs interact with linear peptides bound to cell-surface MHC molecules.

    “Whereas BCRs directly recognize soluble or membrane-bound epitopes, TCRs interact with linear peptides bound to cell-surface major histocompatibility complex (MHC) molecules.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    TCRs and BCRs are transmembrane surface protein complexes that constitute the adaptive immune receptor repertoire.

    “TCRs and BCRs are transmembrane surface protein complexes that constitute the adaptive immune receptor repertoire (AIRR)”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Adding information extracted from histological imaging data can enhance cell identification, resolution of molecular features, and identification of spatial patterns of variation.

    “Adding information extracted from the imaging data can enhance, for example, cell identification193,194 or the resolution of the molecular features195, or can help to identify spatial patterns of variation196.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [193] Tan, X., Su, A., Tran, M. & Nguyen, Q. SpaCell: integrating tissue morphology and spatial gene expression to predict disease cells. Bioinformatics 36, 2293–2294 (2020). source

      Not checked in the free run.

    • cites [194] He, B. et al. Integrating spatial gene expression and breast tumour morphology via deep learning. Nat. Biomed. Eng. 4, 827–834 (2020). source

      Not checked in the free run.

    • cites [195] Bergenstråhle, L. et al. Super-resolved spatial transcriptomics by deep data fusion. Nat. Biotechnol. 40, 476–479 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    BALDR, BASIC and BraCer were shown to robustly recover BCR sequences but are no longer maintained.

    “BALDR169, BASIC170 and BraCer171 were shown to robustly recover BCR sequences172 but are no longer maintained.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [169] Upadhyay, A. A. et al. BALDR: a computational pipeline for paired heavy and light chain immunoglobulin reconstruction in single-cell RNA-seq data. Genome Med. 10, 20 (2018). source

      Not checked in the free run.

    • cites [170] Canzar, S., Neu, K. E., Tang, Q., Wilson, P. C. & Khan, A. A. BASIC: BCR assembly from single cells. Bioinformatics 33, 425–427 (2017). source

      Not checked in the free run.

    • cites [171] Lindeman, I. et al. BraCeR: B-cell-receptor reconstruction and clonality inference from single-cell RNA-seq. Nat. Methods 15, 563–565 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Multigrate learns a joint latent space of paired and unpaired measurements and, combined with transfer learning, can map unimodal and multimodal query data sets to multi-omic references while imputing missing modalities.

    “Alternatively, Multigrate learns a joint latent space of paired and unpaired measurements. Combined with transfer learning, Multigrate can map unimodal and multimodal query data sets to multi-omic references while imputing missing modalities254.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [254] Lotfollahi, M., Litinetskaya, A. & Theis, F. J. Multigrate: single-cell multi-omic data integration. bioRxiv https://doi.org/10.1101/2022.03.16.484643 (2022). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Although no benchmarks exist for TCR sequence reconstruction, MiXCR and TRUST4 are frequently used.

    “Although no benchmarks exist for TCR sequence reconstruction, MiXCR167 and TRUST4 (ref. 168) are frequently used.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [167] Bolotin, D. A. et al. MiXCR: software for comprehensive adaptive immunity profiling. Nat. Methods 12, 380–381 (2015). source

      Not checked in the free run.

    • cites [168] Song, L. et al. TRUST4: immune repertoire reconstruction from bulk and single-cell RNA-seq data. Nat. Methods 18, 627–630 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    All three computational approaches for inferring AIR specificity suffer from reliance on public databases that are biased toward commonly researched diseases and lack MHC information needed to decipher T cell antigen specificity.

    “All three approaches suffer from reliance on public databases that contain data primarily from commonly researched diseases and a lack of information on MHCs to decipher T cell antigen specificity.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Weighted nearest-neighbour analysis learns cell-specific modality weights that reflect modality information content and determine modality importance in downstream analyses as a neighbour graph.

    “Alternatively, weighted nearest-neighbour (WNN)3 analysis learns cell-specific modality weights that reflect the modality information content to determine the importance of modalities in downstream analyses in the form of a neighbour graph.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [3] Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587.e29 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning

    ADT data are less sparse than gene counts, which follow a negative binomial distribution.

    “Contrary to the negative binomial distribution of gene counts, ADT data are less sparse.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Cellular organization bridges the gap between tissue biology and pathology, enabling discovery of new cellular functionalities and creating computational challenges requiring distinct analysis methods.

    “Cellular organization bridges the gap between tissue biology and pathology, which enables the discovery of new cellular functionalities and creates new computational challenges for which distinct analysis methods are required190–192.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [190] Palla, G., Fischer, D. S., Regev, A. & Theis, F. J. Spatial components of molecular tissue biology. Nat. Biotechnol. 40, 308–318 (2022). This review highlights both the importance of different spatial profiling technologies and computational challenges for single-cell data resolved in space. source

      Not checked in the free run.

    • cites [191] Dries, R. et al. Advances in spatial transcriptomic data analysis. Genome Res. 31, 1706–1718 (2021). source

      Not checked in the free run.

    • cites [192] Asp, M., Bergenstråhle, J. & Lundeberg, J. Spatially resolved transcriptomes-next generation tools for tissue exploration. Bioessays 42, e1900221 (2020). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The CDR3 of the VDJ chain is the most influential region for AIR-antigen interaction specificity, with the VJ chain CDR3 having a lesser role.

    “The most influencing positions of the AIR–antigen interaction, reflecting specificity, are contained in the CDR3 of the VDJ chain and to a lesser degree the CDR3 in the VJ chain176.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [176] Springer, I., Tickotsky, N. & Louzoun, Y. Contribution of T cell receptor alpha and beta CDR3, MHC typing, V and J genes to peptide binding prediction. Front. Immunol. 12, 664514 (2021). source

      Not checked in the free run.

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

    Premisecitation

    CiteFuse normalizes ADTs using CLR and combines both modality matrices with a similarity network fusion algorithm.

    “An alternative approach is to use CiteFuse163, which normalizes ADTs using CLR and combines both modality matrices with a similarity network fusion algorithm.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [163] Kim, H. J., Lin, Y., Geddes, T. A., Yang, J. Y. H. & Yang, P. CiteFuse enables multi-modal analysis of CITE-seq data. Bioinformatics 36, 4137–4143 (2020). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    On simulated data sets, cell2location outperformed other cell-type deconvolution approaches but requires more computational resources.

    “For simulated data sets, cell2location outperformed other approaches for cell-type deconvolution, but requires more computational resources, whereas for real data sets, SpatialDWLS and RCTD performed best in terms of the overall accuracy score based on four different accuracy metrics215,216.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [215] Li, B. et al. Benchmarking spatial and single-cell transcriptomics integration methods for transcript distribution prediction and cell type deconvolution. Nat. Methods 19, 662–670 (2022). source

      Not checked in the free run.

    • cites [216] Yan, L. & Sun, X. Benchmarking and integration of methods for deconvoluting spatial transcriptomic data. Bioinformatics 39, btac805 (2023). source

      Not checked in the free run.

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

    Interpretationreasoning

    Processing pipelines such as Giotto and squidpy allow the addition of tailored segmentation methods, simplifying comparison, choice and evaluation of the chosen method.

    “Processing pipelines such as Giotto and squidpy allow the addition of tailored segmentation methods to the analysis pipeline, which simplifies the comparison, choice and evaluation of the chosen method.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationnone given

    Tools for exploratory analysis of patient covariates, integration with omics data sets, and mapping omics measurements to phenotype information are lacking.

    “Tools for their exploratory analysis, the integration with omics data sets and the mapping of omics measurements to phenotype information are lacking, and we expect further developments in this direction.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    Latent representations simplify single-cell data analysis by skipping manual quality control steps.

    “They simplify the analysis of single-cell data by skipping manual quality control steps.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    Extrinsic effects on gene expression variation should be considered when describing cellular organization and tissue niches.

    “These events are commonly referred to as extrinsic effects on gene expression variation and should be taken into consideration in efforts to describe cellular organization and tissue niches236.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [236] Fischer, D. S., Schaar, A. C. & Theis, F. J. Modeling intercellular communication in tissues using spatial graphs of cells. Nat. Biotechnol. https://doi.org/10.1038/s41587-022-01467-z (2022). source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductnone given

    All recommendations in the paper are based on independent benchmarks, which inevitably lag behind the latest method developments.

    “All recommendations are based on independent benchmarks, which inevitably lag behind the latest method developments.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Genes selected for imaging-based spatial transcriptomics based on prior scRNA-seq knowledge may not identify rare cell subpopulations, resulting in bias toward known cell types.

    “Genes are typically selected on the basis of prior biological knowledge obtained from scRNA-seq (probe selection) and might not be suited to the identification of rare cell subpopulations, which results in bias towards known cell types223.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [223] Kuemmerle, L. B. et al. Probe set selection for targeted spatial transcriptomics. Preprint at https://doi.org/10.1101/2022.08.16.504115. source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Coupling genome-scale Perturb-seq with further modalities enables systematic exploration of the genetic landscape to unveil context-specific gene regulatory networks.

    “Coupling genome-scale Perturb-seq with further modalities enables the systematic exploration of the genetic landscape to unveil context-specific gene regulatory networks.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationnone given

    Methods for analysing single-cell proteomics measurements are sparse, selectively benchmarked, and best practices have yet to be developed.

    “Methods for the analysis of these measurements are sparse, selectively benchmarked, and best practices have yet to be developed.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    Gene expression matrices from array-based spatial transcriptomics after deconvolution, or from image-based spatial transcriptomics after segmentation, can be filtered, normalized and visualized similarly to scRNA-seq data.

    “Gene expression matrices obtained by array-based spatial transcriptomics followed by cell-type deconvolution, or by image-based spatial transcriptomics followed by segmentation, can be filtered, normalized and visualized in a similar way to scRNA-seq data.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given

    Methods for identifying spatially variable genes vary broadly in their assumptions and definitions, and there is no consensus on the best approach.

    “Methods for this purpose vary broadly with respect to their assumptions and their definition of SVGs, and there is no consensus on how to best identify SVGs.”
    Quote in the paperSubstantial claim
63 of 63

  • The paper relies on this without support.

    scRNA-seq captures only one layer of the complex regulatory machinery that governs cellular function and signalling.

    “scRNA-seq captures only one layer of the complex regulatory machinery that governs cellular function and signalling.”

  • The paper relies on this without support.

    Heterotypic doublets are hard to annotate and can lead to wrong cell-type labels.

    “Doublets formed by different cell types (heterotypic doublets) are hard to annotate and can lead to wrong cell-type labels.”

  • The authors state this as an assumption.

    Scaling all genes by a common factor assumes that differences in count depth due to cell size are negligible.

    “By scaling all genes by a common factor, one assumes that differences in count depth due to cell size are negligible.”

  • The authors state this as an assumption.

    RNA velocity inference assumes gene independence and constant rates of transcription, splicing and degradation.

    “RNA velocity inference assumes gene independence and constant rates of transcription, splicing and degradation.”

  • The authors state this as an assumption.

    Most analysis tasks assume that each droplet contains RNA from an intact single cell.

    “Most analysis tasks assume that each droplet contains RNA from an intact single cell.”

  • The paper relies on this without support.

    Cell–cell communication is essential for organismal development and homeostasis, and its impairment leads to disease.

    “Cells are in constant interaction with each other for organismal development and homeostasis. If this interaction is impaired, disease ensues.”

  • The paper relies on this without support.

    Not all generated AIR chains produced during allelic rearrangements form a functional AIR.

    “Not all generated AIR chains produced during allelic rearrangements form a functional AIR.”

  • The paper relies on this without support.

    Public clonotypes represent shared immunological responses.

    “Public clonotypes appear in more than one donor and can represent shared immunological response.”

  • The paper relies on this without support.

    Including the VJ CDR3 sequence in database queries decreases the false discovery rate compared with using the VDJ CDR3 sequence alone.

    “Similarly to clonotype assignment, database queries can be conducted with varying strictness by considering either the VDJ CDR3 sequence alone, or additionally the VJ CDR3 sequence, which decreases the FDR.”

  • The paper relies on this without support.

    The presence of AIRs can guide cell-type annotation by separating immune cell clusters and facilitating detailed T cell annotations.

    “The presence of AIRs can guide cell-type annotation by separating immune cell clusters and facilitating detailed T cell annotations.”

  • The paper relies on this without support.

    Analysing spatial data sets requires analysis tools specifically tailored to this modality.

    “Analysing spatial data sets requires analysis tools specifically tailored to this modality, which can be conducted with frameworks such as Squidpy209, Giotto210, Seurat45 or SpatialExperiment211.”

  • The paper relies on this without support.

    In tissue, cells interact through direct contact, surface-bound ligands and receptors, long-range paracrine effects, bio-mechanical forces, and indirect mechanisms such as metabolite exchange.

    “In tissue, cells have direct contact and can interact through surface-bound ligands and receptors, long-range paracrine effects, bio-mechanical forces and indirect mechanisms such as metabolite exchange.”

  • The paper relies on this without support.

    Array-based spatial transcriptomics assays do not capture single-cell resolution, so spot gene expression reflects cell-type composition rather than distinct cell types.

    “As array-based assays do not capture single-cell resolution, the gene expression profile of spots reflects cell-type composition rather than distinct cell types.”

  • The paper relies on this without support.

    Imaging-based spatial transcriptomics technologies commonly read out only a predefined set of transcripts.

    “These technologies commonly read out only a predefined set of transcripts.”

  • The paper relies on this without support.

    Alignment of spatially naive scRNA-seq data and targeted spatially resolved data enables imputation of the whole transcriptome in a spatially resolved manner.

    “Alignment of standard spatially naive scRNA-seq data and targeted spatially resolved data enables imputation of the whole transcriptome (measured in standard scRNA-seq) in a spatially resolved manner and attempts to resolve the limitations of targeted feature spaces.”

  • The paper relies on this without support.

    The identification of cellular microenvironments across different samples can be hindered by differences in image orientation.

    “The identification of cellular microenvironments across different samples can be hindered by differences with respect to image orientation.”

  • The paper relies on this without support.

    Cell communication events can be identified in dissociated scRNA-seq data.

    “Cell communication events can be identified in dissociated scRNA-seq data as described above.”

  • The paper relies on this without support.

    Models built on latent spaces become predictive with query-to-reference mapping approaches, creating a shift from unsupervised exploratory analysis to supervised predictions.

    “Models built on these latent spaces become predictive with query-to-reference mapping approaches, which will create a shift from the unsupervised, exploratory analysis approach to single-cell analysis complemented by supervised predictions.”
18 of 18

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?

  • There is no consensus on which method represents the state of the art for low-MOI association testing; these methods have not undergone rigorous statistical validation and comparison.

    “there is no consensus as to which of these methods represents the “state of the art”; these methods have not undergone rigorous statistical validation and comparison [18]”

    From Robust differential expression testing for single-cell CRISPR screens at low multiplicity of infection (2024) · cites this paper as [18]

    In Transcriptome > Revealing mechanisms (page unavailable), the paper says that perturbation-modelling tools have not yet been independently benchmarked, supporting the concern about method comparison. It does not specifically address low-MOI association testing, identify a lack of consensus on its state-of-the-art method, or assess those methods’ statistical validation. (LLM, conf=0.82)

    Experimental — accuracy not yet measured

  • Development of single-cell multi-omics methods and growth of single-cell datasets created a well-suited scenario for applying deep learning methods in single-cell analysis, which could further lead to advancements in drug discovery, therapeutic targets, and digital health.

    “The development of single-cell multi-omics methods and the growth of single-cell datasets created a well-suited scenario for applying deep learning methods in single-cell analysis, which could further lead to advancements in the field of drug discovery, therapeutic targets, and digital health [2,131].”

    From Challenges in AI-driven Biomedical Multimodal Data Fusion and Analysis (2025) · cites this paper as [2]

    Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

    Experimental — accuracy not yet measured

  • Corrected count matrices for each sample were processed using Scanpy tools following recommended best practices for retaining high-quality cells.

    “Corrected count matrices for each sample were processed using Scanpy tools following recommended best practices for retaining high-quality cells63.”

    From A comparative transcriptomic analysis of mouse demyelination models and multiple sclerosis lesions (2026) · cites this paper as [63]

  • There is no commonly accepted framework for quantifying the statistical validity of single-cell CRISPR screen association testing methods.

    “there is no commonly accepted framework for quantifying the statistical validity of single-cell CRISPR screen association testing methods”

    From Robust differential expression testing for single-cell CRISPR screens at low multiplicity of infection (2024) · cites this paper as [18]

    The Introduction says that a lack of best-practice workflows makes single-cell tools difficult to navigate. The Transcriptome > Revealing mechanisms section discusses perturbation-modelling methods and statistical validity in differential gene-expression testing, but does not address whether a commonly accepted framework exists for assessing the statistical validity of single-cell CRISPR screen association tests. Searched for CRISPR association testing, statistical validity, and benchmarking frameworks; no passage in our copy settles the claim (page numbers unavailable). (LLM, conf=0.73)

    Experimental — accuracy not yet measured

  • A common task in single-cell analysis is to eliminate batch effects without removing meaningful biological differences, allowing for data integration.

    “Hence, a common task in single-cell analysis is to eliminate batch effects without removing meaningful biological differences, allowing for data integration [13, 14, 15].”

    From Assessing the limits of zero-shot foundation models in single-cell biology (2023) · cites this paper as [15]

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.

  • Recommendation claim overlooks unbenchmarked guidance

    Minor

    The conclusion says every recommendation rests on an independent benchmark, but the paper also advises an analysis step while acknowledging the absence of a systematic benchmark for it.

1 of 1

Cites

Flagged first, then by citations

6 of the 269 references couldn’t be matched to a known paper.

Research works printed but not linked: 2

  1. Ahlmann-Eltze, C. & Huber, W. Comparison of transformations for single-cell RNA-seq data. bioRxiv https://doi.org/10.1101/2021.06.24.449781 (2022). · DOI · Paper page
  2. Sokal, R. & Michener, C. A statistical method for evaluating systematic relationships. Univ. Kans., Sci. Bull. 38, 1409–1438 (1958).

Cited by

Sorted by citations

Related

Authors
Heumos, L., Schaar, A. C., Lance, C., Litinetskaya, A., Drost, F., Zappia, L., Lücken, M. D., Strobl, D. C., Henao, J., Curion, F., Single-cell Best Practices Consortium, Aliee, H., Ansari, M., Badia-i-Mompel, P., Büttner, M., Dann, E., Dimitrov, D., Dony, L., Frishberg, A., He, D., Hediyeh-zadeh, S., Hetzel, L., Ibarra, I. L., Jones, M. G., Lotfollahi, M., Martens, L. D., Müller, C. L., Nitzan, M., Ostner, J., Palla, G., Patro, R., Piran, Z., Ramírez-Suástegui, C., Saez-Rodriguez, J., Sarkar, H., Schubert, B., Sikkema, L., Srivastava, A., Tanevski, J., Virshup, I., Weiler, P., Schiller, H. B., & Theis, F. J.
Affiliations
  • Helmholtz Munich (DE)
  • Heidelberg University (DE)
  • University of Bonn (DE)
  • Wellcome Sanger Institute (GB)
  • University of Maryland, College Park (US)
  • Stanford Medicine (US)
  • Hebrew University of Jerusalem (IL)
  • La Jolla Institute for Immunology (US)
  • Harvard University (US)
  • New York Genome Center (US)
Corresponding author
Theis, Fabian J.
ORCID
Venue
Nature Reviews Genetics
Publisher
Springer Science and Business Media LLC
Published
2023-03-31
Volume
24
Issue
8
Pages
550-572
ISSN
  • 1471-0056 (ISSN-L)
  • 1471-0064
DOI
10.1038/s41576-023-00586-w
PMC
PMC10066026
OpenAlex
W4362521490
Cited by
1,051
References
269
Citations by year
2026: 295 · 2025: 367 · 2024: 296 · 2023: 90 · 2022: 2
FWCI
28.10
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Reviews Genetics (publishedVersion)
  • PubMed Central (publishedVersion)
  • PubMed (publishedVersion)
Funders
  • National Science Foundation (1763680, 1750472)
  • Joachim Herz Stiftung
  • Wellcome Trust (108413 /A/15/D)
  • Deutsche Forschungsgemeinschaft (452881907, SFB-TRR 338/1, TRR355, TP Z02, SFB-TRR 338, 490846870)
  • Bundesministerium für Bildung und Forschung (01IS18053A)
  • Helmholtz Artificial Intelligence Cooperation Unit (ZT-I-PF-5-01)
  • National Institutes of Health
  • National Human Genome Research Institute (R01HG009937)
  • National Cancer Institute (K99CA267677)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
Keywords
single-cell multi-omics, multimodal data integration, single-cell transcriptomics, chromatin accessibility, surface protein expression, adaptive immune receptor repertoire
MeSH terms
Gene Expression Profiling, Proteomics, Single-Cell Analysis

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 2026-10-07.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-07.

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