CoChat

Resolving cellular systems by ultra-sensitive and economical single-cell transcriptome filtering

Open access

Preprint — published as 10.1016/j.isci.2021.102147

The DOI matches its Crossref record; 27 of the 27 references it cites were found in a registry or an index.

Abstract

Single-cell transcriptomics has sensitivity limits that restrict low abundance transcript identification, affects clustering and introduce artefact. Here, we describe Constellation DropSeq (C-DropSeq), a molecular transcriptome filter that delivers two orders of magnitude sensitivity gains by maximising read utility while reducing sequencing depth and costs. The simple and powerful method is broadly compatible with library preparation routines and was demonstrated by identifying and characterizing the activation of rare dendritic cell sub-populations.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1101/800631

Cites

27 of 27 references verified0 citation errors · 0 not checkable

Statistics

Figures

Graphical abstract
Graphical abstract CC BY · PubMed Central
Figure 1 Constellation-Seq methodology and performance (A) Schematic representation of the method: Constellation-Seq can be applied to any Smart-Seq-like library following the standard cDNA synthesis protocol. With a defined primer panel, Constellation-Seq can be applied directly to the cDNA library (black arrows). Otherwise, an aliquot of the cDNA can be used for bulk sequencing and after data analysis the panel of primers can be selected for hypothesis testing or to reduce the technical zeros (Blue arrows). Constellation-Seq includes a hybrid primer (14–18 bp specific sequence, black, adjacent to a common 14 bp handle 2, red) that binds to a specific target sequence in the cDNA library. Linear amplification of 500–1000 bp stretches of target transcripts allows selective enrichment of targets of interest, and the inclusion of the cell barcode and UMI sequences, leads to generation of the Constellation library, ready to use in next-generation sequencing. (B-E) Constellation-Seq was compared against standard sequencing using a panel of 52 targets on control beads. (B) UMAP representation of control beads with standard sequencing compared with Constellation-Seq. (C) A track plot showing the reduction in the data sparsity in a head to head comparison. Each bar represents a gene expression signal from a single cell. A full track plot is included as Figure S4. (D) Individual target raw counts show ~100-fold sensitivity gains for Constellation-Seq, error bars represent SD. (E) Dramatic reduction in technical zeros achieved by Constellation-Seq compared with DropSeq. At 2K UMI counts per bead 32/49 genes were detected in half of the beads (3 negative controls were not detected) using Constellation-Seq, whereas only 1 was detected with the same threshold using DropSeq.
Figure 1 Constellation-Seq methodology and performance (A) Schematic representation of the method: Constellation-Seq can be applied to any Smart-Seq-like library following the standard cDNA synthesis protocol. With a defined primer panel, Constellation-Seq can be applied directly to the cDNA library (black arrows). Otherwise, an aliquot of the cDNA can be used for bulk sequencing and after data analysis the panel of primers can be selected for hypothesis testing or to reduce the technical zeros (Blue arrows). Constellation-Seq includes a hybrid primer (14–18 bp specific sequence, black, adjacent to a common 14 bp handle 2, red) that binds to a specific target sequence in the cDNA library. Linear amplification of 500–1000 bp stretches of target transcripts allows selective enrichment of targets of interest, and the inclusion of the cell barcode and UMI sequences, leads to generation of the Constellation library, ready to use in next-generation sequencing. (B-E) Constellation-Seq was compared against standard sequencing using a panel of 52 targets on control beads. (B) UMAP representation of control beads with standard sequencing compared with Constellation-Seq. (C) A track plot showing the reduction in the data sparsity in a head to head comparison. Each bar represents a gene expression signal from a single cell. A full track plot is included as Figure S4. (D) Individual target raw counts show ~100-fold sensitivity gains for Constellation-Seq, error bars represent SD. (E) Dramatic reduction in technical zeros achieved by Constellation-Seq compared with DropSeq. At 2K UMI counts per bead 32/49 genes were detected in half of the beads (3 negative controls were not detected) using Constellation-Seq, whereas only 1 was detected with the same threshold using DropSeq.CC BY · PubMed Central
Figure 2 Constellation-Seq reliably measures gene expression changes with higher sensitivity (A) Experimental design; PBMCs from healthy subjects (n = 3) were stimulated with Staphylococcal enterotoxin B (SEB) or media control for 16hr and analyzed using DropSeq and Constellation-Seq. (B) Correlation of normalized gene expression fold changes induced by SEB as detected by DropSeq and Constellation-Seq. Pseudo-bulk counts for each gene used for the comparison. (C) Comparative analysis of selected markers induced by SEB in cultured PBMCs. Violin plots in each row show the distribution and levels of each expressed gene in different culture conditions (CTR – media control, SEB – stimulated cells) and assessed by DropSeq (gray) and Constellation-Seq (orange). y axis represents normalized UMI counts.
Figure 2 Constellation-Seq reliably measures gene expression changes with higher sensitivity (A) Experimental design; PBMCs from healthy subjects (n = 3) were stimulated with Staphylococcal enterotoxin B (SEB) or media control for 16hr and analyzed using DropSeq and Constellation-Seq. (B) Correlation of normalized gene expression fold changes induced by SEB as detected by DropSeq and Constellation-Seq. Pseudo-bulk counts for each gene used for the comparison. (C) Comparative analysis of selected markers induced by SEB in cultured PBMCs. Violin plots in each row show the distribution and levels of each expressed gene in different culture conditions (CTR – media control, SEB – stimulated cells) and assessed by DropSeq (gray) and Constellation-Seq (orange). y axis represents normalized UMI counts.CC BY · PubMed Central
Figure 3 Constellation-Seq is compatible with standard 10-X protocol (A) DropSeq assay of human monocytes (A) The spikes in the plot are characteristic for the Constellation-Seq method due to the selection of targets with distinct molecular weights. (B) Direct comparison of gene expression in monocytes using Constellation-Seq vs 10X. (C) Left: Constellation-Seq (orange) vs 10X (gray), gene expression measured in normalized UMI counts (C). UMAP plot of 6,000 monocyte transcriptomes assessed using 10X and C-10X. At 1.5K UMI counts per cell, C-10X shows more granularity than normal 10X at the same resolution. The enhanced sensitivity of C-10X is demonstrated using monocyte markers.
Figure 3 Constellation-Seq is compatible with standard 10-X protocol (A) DropSeq assay of human monocytes (A) The spikes in the plot are characteristic for the Constellation-Seq method due to the selection of targets with distinct molecular weights. (B) Direct comparison of gene expression in monocytes using Constellation-Seq vs 10X. (C) Left: Constellation-Seq (orange) vs 10X (gray), gene expression measured in normalized UMI counts (C). UMAP plot of 6,000 monocyte transcriptomes assessed using 10X and C-10X. At 1.5K UMI counts per cell, C-10X shows more granularity than normal 10X at the same resolution. The enhanced sensitivity of C-10X is demonstrated using monocyte markers.CC BY · PubMed Central
Figure 4 Constellation-Seq can resolve rare cell populations 10X Chromium Single Cell 3′ assay of human PBMC. (A) UMAP projection of 4,182 PBMCs (Leiden r = 0.5, n_pcs = 20, n_neighbours = 20). Cells were grouped into eight clusters. Classification of PBMCs was inferred from the annotation of cluster-specific genes and based on expression of well-known markers of immune cell types using all detected genes. (B) Sub-clustering of dendritic cells (n = 51), from the standard 10X protocol showing marker genes for DC cell populations as in the Villani paper (Villani et al., 2017). (C) Constellation-Seq run of the sample with a panel of 127 genes related to DC biology (n = 1,697 cells). (D) UMAP plot of the dendritic cell subsets (n = 127).
Figure 4 Constellation-Seq can resolve rare cell populations 10X Chromium Single Cell 3′ assay of human PBMC. (A) UMAP projection of 4,182 PBMCs (Leiden r = 0.5, n_pcs = 20, n_neighbours = 20). Cells were grouped into eight clusters. Classification of PBMCs was inferred from the annotation of cluster-specific genes and based on expression of well-known markers of immune cell types using all detected genes. (B) Sub-clustering of dendritic cells (n = 51), from the standard 10X protocol showing marker genes for DC cell populations as in the Villani paper (Villani et al., 2017). (C) Constellation-Seq run of the sample with a panel of 127 genes related to DC biology (n = 1,697 cells). (D) UMAP plot of the dendritic cell subsets (n = 127).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

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.

The self-check needs the paper’s full text.

Citations

Cites

Show all 27 references

Cited by

Metadata

Authors
Vallejo, A. F., Davies, J., Grover, A., Tsai, C.-H., Jepras, R., Polak, M. E., & West, J.
Affiliations
  • University of Southampton (GB)
  • GlaxoSmithKline (United Kingdom) (GB)
  • GlaxoSmithKline (United States) (US)
Corresponding author
  • Polak, Marta E.
  • West, Jonathan
Venue
iScience
Publisher
openRxiv
Published
2019-10-10
Volume
24
Issue
3
Pages
102147-102147
ISSN
2589-0042 (ISSN-L)
OpenAlex
W3128091223
Cited by
17
References
26
Citations by year
2026: 2 · 2025: 2 · 2023: 2 · 2022: 4 · 2021: 2 · 2020: 5
FWCI
0.63
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • iScience (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • ePrints Soton (University of Southampton) (acceptedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • ePrints Soton (University of Southampton) (submittedVersion)
Funders
  • GlaxoSmithKline
  • Wellcome Trust (109377/Z/15/Z, /Z/15/Z, 109377)
  • University of Southampton
  • Medical Research Council (MC_PC_15078)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
Keywords
single-cell transcriptomics, sequencing sensitivity, sequencing cost reduction, dendritic cells, peripheral blood mononuclear cells, gene expression profiling, marker validation

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.

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

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.