A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors

doi:10.1038/s41591-020-0844-1

. 2020 May;26(5):792-802.

doi: 10.1038/s41591-020-0844-1. Epub 2020 May 11.

A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors

Michal Slyper^#¹, Caroline B M Porter^#¹, Orr Ashenberg^#¹, Julia Waldman¹, Eugene Drokhlyansky¹, Isaac Wakiro^{2

3

4}, Christopher Smillie¹, Gabriela Smith-Rosario¹, Jingyi Wu^{2

3

4}, Danielle Dionne¹, Sébastien Vigneau^{2

3

4}, Judit Jané-Valbuena¹, Timothy L Tickle¹, Sara Napolitano^{2

3

4}, Mei-Ju Su^{2

3

4}, Anand G Patel^{5

6}, Asa Karlstrom⁵, Simon Gritsch^{7

8}, Masashi Nomura^{7

8}, Avinash Waghray⁹, Satyen H Gohil^{2

3}, Alexander M Tsankov^{1

10}, Livnat Jerby-Arnon¹, Ofir Cohen^{2

3

4}, Johanna Klughammer¹, Yanay Rosen¹, Joshua Gould¹, Lan Nguyen¹, Matan Hofree¹, Peter J Tramontozzi¹¹, Bo Li^{2

12}, Catherine J Wu^{2

3

13

14}, Benjamin Izar^{2

3

4

15

16

17

18

19}, Rizwan Haq^{2

3}, F Stephen Hodi^{3

18}, Charles H Yoon^{3

20}, Aaron N Hata^{8

21}, Suzanne J Baker⁵, Mario L Suvà^{2

7

8}, Raphael Bueno¹¹, Elizabeth H Stover^{2

3}, Michael R Clay²², Michael A Dyer^{5

23}, Natalie B Collins^{2

24

25}, Ursula A Matulonis³, Nikhil Wagle^{2

3

4

13

14}, Bruce E Johnson^{3

4}, Asaf Rotem^{2

3

4}, Orit Rozenblatt-Rosen²⁶, Aviv Regev^{27

28

29}

Affiliations

¹ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA.
² Broad Institute of Harvard and MIT, Cambridge, MA, USA.
³ Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA.
⁴ Center for Cancer Precision Medicine of Dana-Farber Cancer Institute, Boston, MA, USA.
⁵ Department of Developmental Neurobiology, St Jude Children's Research Hospital, Memphis, TN, USA.
⁶ Department of Oncology, St Jude Children's Research Hospital, Memphis, TN, USA.
⁷ Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
⁸ Center for Cancer Research, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
⁹ Center for Regenerative Medicine, Massachusetts General Hospital, Boston, MA, USA.
¹⁰ Icahn School of Medicine at Mount Sinai, New York, NY, USA.
¹¹ Division of Thoracic Surgery, Brigham and Women's Hospital, Boston, MA, USA.
¹² Center for Immunology and Inflammatory Diseases, Division of Rheumatology, Allergy, and Immunology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
¹³ Department of Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA.
¹⁴ Harvard Medical School, Boston, MA, USA.
¹⁵ Laboratory for Systems Pharmacology, Harvard Medical School, Boston, MA, USA.
¹⁶ Center for Immunology and Virology, Dana-Farber Cancer Institute, Boston, MA, USA.
¹⁷ Ludwig Center for Cancer Research at Harvard, Boston, MA, USA.
¹⁸ Melanoma Disease Center, Dana-Farber Cancer Institute, Boston, MA, USA.
¹⁹ Columbia Center for Translational Immunology and Division of Hematology and Oncology, Columbia University Medical Center, New York, NY, USA.
²⁰ Department of Surgical Oncology, Brigham and Women's Hospital, Boston, MA, USA.
²¹ Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
²² Department of Pathology, St Jude Children's Research Hospital, Memphis, TN, USA.
²³ Howard Hughes Medical Institute, Chevy Chase, MD, USA.
²⁴ Department of Pediatric Oncology, Dana-Farber Cancer Institute, Boston, MA, USA.
²⁵ Division of Pediatric Hematology and Oncology, Boston Children's Hospital, Boston, MA, USA.
²⁶ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA. orit@broadinstitute.org.
²⁷ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA. aregev@broadinstitute.org.
²⁸ Howard Hughes Medical Institute, Chevy Chase, MD, USA. aregev@broadinstitute.org.
²⁹ Koch Institute for Integrative Cancer Research, Department of Biology, Massachusetts Institute of Technology, Cambridge, MA, USA. aregev@broadinstitute.org.

^# Contributed equally.

PMID: 32405060
PMCID: PMC7220853
DOI: 10.1038/s41591-020-0844-1

A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors

Michal Slyper et al. Nat Med. 2020 May.

. 2020 May;26(5):792-802.

doi: 10.1038/s41591-020-0844-1. Epub 2020 May 11.

Authors

Affiliations

¹ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA.
² Broad Institute of Harvard and MIT, Cambridge, MA, USA.
³ Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA.
⁴ Center for Cancer Precision Medicine of Dana-Farber Cancer Institute, Boston, MA, USA.
⁵ Department of Developmental Neurobiology, St Jude Children's Research Hospital, Memphis, TN, USA.
⁶ Department of Oncology, St Jude Children's Research Hospital, Memphis, TN, USA.
⁷ Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
⁸ Center for Cancer Research, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
⁹ Center for Regenerative Medicine, Massachusetts General Hospital, Boston, MA, USA.
¹⁰ Icahn School of Medicine at Mount Sinai, New York, NY, USA.
¹¹ Division of Thoracic Surgery, Brigham and Women's Hospital, Boston, MA, USA.
¹² Center for Immunology and Inflammatory Diseases, Division of Rheumatology, Allergy, and Immunology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
¹³ Department of Internal Medicine, Brigham and Women's Hospital, Boston, MA, USA.
¹⁴ Harvard Medical School, Boston, MA, USA.
¹⁵ Laboratory for Systems Pharmacology, Harvard Medical School, Boston, MA, USA.
¹⁶ Center for Immunology and Virology, Dana-Farber Cancer Institute, Boston, MA, USA.
¹⁷ Ludwig Center for Cancer Research at Harvard, Boston, MA, USA.
¹⁸ Melanoma Disease Center, Dana-Farber Cancer Institute, Boston, MA, USA.
¹⁹ Columbia Center for Translational Immunology and Division of Hematology and Oncology, Columbia University Medical Center, New York, NY, USA.
²⁰ Department of Surgical Oncology, Brigham and Women's Hospital, Boston, MA, USA.
²¹ Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
²² Department of Pathology, St Jude Children's Research Hospital, Memphis, TN, USA.
²³ Howard Hughes Medical Institute, Chevy Chase, MD, USA.
²⁴ Department of Pediatric Oncology, Dana-Farber Cancer Institute, Boston, MA, USA.
²⁵ Division of Pediatric Hematology and Oncology, Boston Children's Hospital, Boston, MA, USA.
²⁶ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA. orit@broadinstitute.org.
²⁷ Klarman Cell Observatory, Broad Institute of Harvard and MIT, Cambridge, MA, USA. aregev@broadinstitute.org.
²⁸ Howard Hughes Medical Institute, Chevy Chase, MD, USA. aregev@broadinstitute.org.
²⁹ Koch Institute for Integrative Cancer Research, Department of Biology, Massachusetts Institute of Technology, Cambridge, MA, USA. aregev@broadinstitute.org.

^# Contributed equally.

PMID: 32405060
PMCID: PMC7220853
DOI: 10.1038/s41591-020-0844-1

Erratum in

Author Correction: A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors.
Slyper M, Porter CBM, Ashenberg O, Waldman J, Drokhlyansky E, Wakiro I, Smillie C, Smith-Rosario G, Wu J, Dionne D, Vigneau S, Jané-Valbuena J, Tickle TL, Napolitano S, Su MJ, Patel AG, Karlstrom A, Gritsch S, Nomura M, Waghray A, Gohil SH, Tsankov AM, Jerby-Arnon L, Cohen O, Klughammer J, Rosen Y, Gould J, Nguyen L, Hofree M, Tramontozzi PJ, Li B, Wu CJ, Izar B, Haq R, Hodi FS, Yoon CH, Hata AN, Baker SJ, Suvà ML, Bueno R, Stover EH, Clay MR, Dyer MA, Collins NB, Matulonis UA, Wagle N, Johnson BE, Rotem A, Rozenblatt-Rosen O, Regev A. Slyper M, et al. Nat Med. 2020 Aug;26(8):1307. doi: 10.1038/s41591-020-0976-3. Nat Med. 2020. PMID: 32587393 Free PMC article.

Abstract

Single-cell genomics is essential to chart tumor ecosystems. Although single-cell RNA-Seq (scRNA-Seq) profiles RNA from cells dissociated from fresh tumors, single-nucleus RNA-Seq (snRNA-Seq) is needed to profile frozen or hard-to-dissociate tumors. Each requires customization to different tissue and tumor types, posing a barrier to adoption. Here, we have developed a systematic toolbox for profiling fresh and frozen clinical tumor samples using scRNA-Seq and snRNA-Seq, respectively. We analyzed 216,490 cells and nuclei from 40 samples across 23 specimens spanning eight tumor types of varying tissue and sample characteristics. We evaluated protocols by cell and nucleus quality, recovery rate and cellular composition. scRNA-Seq and snRNA-Seq from matched samples recovered the same cell types, but at different proportions. Our work provides guidance for studies in a broad range of tumors, including criteria for testing and selecting methods from the toolbox for other tumors, thus paving the way for charting tumor atlases.

PubMed Disclaimer

Conflict of interest statement

A. Regev is a founder of and equity holder in Celsius Therapeutics, an equity holder in Imunitas, and a SAB member of Syros Pharmaceuticals, Thermo Fisher Scientific, Asimov and NeoGene Therapeutics. M.S., O.A., E.D., O.R.-R. and A. Regev are co-inventors on patent applications filed by the Broad Institute for inventions relating to work in this manuscript, such as in PCT/US2018/060860 and US provisional application no. 62/745,259. C.J.W. is a founder and member of the scientific advisory board of Neon Therapeutics and receives research funding from Pharmacyclics. A. Rotem is a consultant and equity holder in Celsius Therapeutics. F.S.H. reports grants, personal fees from Bristol-Myers Squibb and Novartis and personal fees from Merck, EMD Serono, Takeda, Surface, Genentech/Roche, Compass Therapeutics, Apricity, Bayer, Aduro, Sanofi, Pfizer, Pionyr, Verastem, Torque and Rheos. In addition, F.S.H. has a patent ‘Methods for Treating MICA-Related Disorders’ (#20100111973) with royalties paid, a patent ‘Tumor Antigens and Uses Thereof’ (#7250291) issued, a patent ‘Angiopoiten-2 Biomarkers Predictive of Anti-immune Checkpoint Response’ (#20170248603) pending, a patent ‘Compositions and Methods for Identification, Assessment, Prevention and Treatment of Melanoma using PD-L1 Isoforms’ (#20160340407) pending, a patent ‘Therapeutic Peptides’ (#20160046716) pending, a patent ‘Therapeutic Peptides’ (#20140004112) pending, a patent ‘Therapeutic Peptides’ (#20170022275) pending, a patent ‘Therapeutic Peptides’ (#20170008962) pending, a patent ‘Therapeutic Peptides’ (patent no. 9402905) issued and a patent Methods of using ‘Pembrolizumab and Trebananib’ pending. R.H. has received research support from Novartis and Bristol-Myers and is a consultant for Tango Therapeutics. R.B. has received research support from Roche, Genentech, Merck, Siemens, Verastem, Gritstone, Epizyme, Medgenome and HTG and has equity in Navigation Sciences. A.N.H. has received research support from Novartis, Amgen, Pfizer, Roche/Genentech and Relay Therapeutics. B.I. is a paid consultant for Merck. U.A.M. received consulting fees from Novartis and Merck.

Figures

**Fig. 1. Study and toolbox overview.**
a, sc/snRNA-Seq workflow, experimental and computational pipelines, and protocol selection criteria. b, Tumor types and samples processed in the study. Tested and selected protocols for fresh (white circles, cells), frozen (blue circles, nuclei) and cryopreserved (underlined circles, cells (white) and nuclei (blue)) are indicated. O-PDX, orthotopic patient-derived xenograft; EZ, EZPrep; ST, salts and Tris; QC, quality control; DE, differentially expressed; InferCNV, Infer Copy Number Variation, a method for detecting copy number abberations; C4, collagenase 4 and DNase I; PDEC, pronase, dispase, elastase, collagenases A and 4 and DNase I; LE, Liberase TM, elastase and DNase I; EZ, EZPrep; NST, Nonidet P40 with salts and Tris; CST, CHAPS with salts and Tris; TST, Tween with salts and Tris; LD, Liberase TM and DNase I; BTD, brain tumor dissociation; MHTD, Miltenyi Biotec human tumor dissociation.

**Fig. 2. Fresh tumor processing and protocol selection for scRNA-Seq.**
a, Flow chart recommended for collection and processing of fresh tumor samples. b–f, Comparison of three dissociation protocols applied to one NSCLC sample. b, Protocol performance varies across cell types. Top and middle: distribution (median and first and third quartiles) of the number of reads per cell, the number of UMIs per cell, the number of genes per cell and fraction of UMIs per cell mapping to mitochondrial genes (Fr. mito. genes) (y axes) in each protocol (x axis) across the entire dataset. Bottom: distribution (median and first and third quartiles) of the number of genes per cell (y axis) only in epithelial cells (left) or in B cells (right). c, The protocols detect similar numbers of doublets. Uniform manifold approximation and projection (UMAP) embedding of single cell profiles (dots) for each protocol, colored by assignment as single cell (gray) or doublet (red). Horizontal bars (bottom): fraction of single (gray) and doublet (red) cells. d, The protocols vary in the number of empty drops. UMAP embedding of single cell profiles (dots) for each protocol, colored by assignment as cell (gray) or empty drop (red). Horizontal bars (bottom): fraction of assigned cells (gray) and empty drops (red). e, The protocols vary in the diversity of cell types captured. UMAP embedding of single cell profiles (dots) from all three protocols, colored by assigned cell subset signature (left) or by protocol (right). Bottom: proportion of cells in each subset in each of the three protocols; k, number of cells passing QC. f, Inferred CNA profiles. Chromosomal amplification (red) and deletion (blue) are inferred in each chromosomal position (columns) across the single cells (rows) using the PDEC protocol. Top: reference cells not expected to contain CNAs in this tumor. Bottom: cells tested for CNAs relative to the reference cells. Color bar: assigned cell type signature for each cell. g, Successful depletion of CD45⁺ cells. The proportion of cells in each subset without and with CD45⁺ depletion in NSCLCs (top) and ovarian ascites (bottom) samples is shown; k, number of cells passing QC. n = 1 sample per protocol. The numbers of cells (k) are indicated in e and g. Numbers of epithelial cells from NSCLC-C4, PDEC and LE are k = 1,284, 641 and 260, respectively, and the number of B cells is k = 100, 121 and 78, respectively. ACK, ammonium-chloride-potassium.

**Fig. 3. scRNA-Seq protocol comparison across tumor types.**
a–d, QC metrics. The number of UMIs per cell (a), number of genes per cell (b), fraction of UMIs per cell mapping to mitochondrial genes (c) and fraction of empty drops (d) (x axes) for each sample (y axis). Median and first and third quartiles are shown in a–c. e, Cell type composition. Proportion of cells assigned to each cell type signature (color) for each sample. O-PDX, orthotopic patient-derived xenograft. Tested protocols for processing each tumor type are indicated. f, Inferred CNA profiles for matched pre- and post-treatment neuroblastoma samples. Chromosomal amplification (red) and deletion (blue) inferred in each chromosomal position (columns) across the single cells (rows) from pre-treatment biopsy HTAPP-312-SMP-901 (left) and post-treatment resection HTAPP-312-SMP-902 (right). Top: reference cells not expected to contain CNAs in this tumor. Bottom: cells tested for CNAs relative to the reference cells. Color bars: assigned cell type signature for each cell. n = 1 sample per protocol. The numbers of cells (k) are indicated in e. Prot., protocol; NB, neuroblastoma; OPC, oligodendrocyte progenitor like cell; NPC, neural progenitor like cell; NK, natural killer; k., kit.

**Fig. 4. Frozen tumor processing and protocol selection for snRNA-Seq.**
a, Flow chart for collection and processing of frozen tumor samples. b–e, Comparison of four nucleus isolation protocols in one neuroblastoma sample. b, Variation in protocol performance: distributions (median and first and third quartiles) of the number of UMIs per nucleus, the number of genes per nucleus and fraction of UMIs per nucleus mapping to mitochondrial genes (y axes) in each protocol (x axis) across all nuclei in the dataset. c, The protocols detect similar numbers of doublets. UMAP embedding of single nucleus profiles (dots) for each protocol is colored by assignment as nucleus (gray) or doublet (red). Horizontal bars (bottom): fraction of single (gray) and doublet (red) nuclei. d, The protocols vary in the diversity of cell types captured. UMAP embedding of single nucleus profiles (dots) from all four protocols is colored by assigned cell subset signature (left) or protocol (right). Bottom: proportion of cells from each subset in each of the protocols. k, number of nuclei passing QC. e, Inferred CNA profiles. Chromosomal amplification (red) and deletion (blue) inferred in each chromosomal position (columns) across the single nuclei (rows) from the TST protocol are shown. Top: reference nuclei not expected to contain CNAs in this tumor. Bottom: nuclei tested for CNAs relative to the reference nuclei. Color bar: assigned cell type signature for each nucleus. n = 1 sample per protocol. The numbers of nuclei (k) are indicated in d. OCT, optimal cutting temperature compound.

**Fig. 5. snRNA-Seq protocol comparison across tumor types.**
a–c, QC metrics: distributions (median and first and third quartiles) of the number of UMIs per nucleus (a), the number of genes per nucleus (b) and the fraction of UMIs per nucleus mapping to mitochondrial genes (c) (x axes) for each sample (y axis). d, Cell type composition, showing the proportion of nuclei assigned to each cell type signature (color) for each sample. n = 1 sample per protocol. The numbers of nuclei (k) are indicated in d.

**Fig. 6. scRNA-Seq and snRNA-Seq recover comparable cells in different proportions.**
a–g, Neuroblastoma. a–c, UMAP embedding of scRNA-Seq and snRNA-Seq profiles of the same neuroblastoma sample combined by CCA (Methods) showing profiles (dots) from both (a), scRNA-Seq (b) and snRNA-Seq (c), colored by the assigned cell type signatures. d, Proportion of cells from each subset in the two protocols. k, number of cells or nuclei passing QC. e,f, Same UMAP embedding as in a, colored by cells or nuclei (e) or by unsupervised clustering (f). g, Fraction of cells and nuclei in each cluster from f. n = 1 sample per protocol. The numbers of cells and nuclei (k) are indicated in d. h–n, MBC. As in a–g for an MBC sample. n = 1 sample per protocol. The numbers of cells and nuclei are indicated in k.

**Extended Data Fig. 1. scRNA-Seq protocol comparison for a single NSCLC sample.**
**(a)** Sample processing and QC overview. For each protocol, shown are the number of cells passing QC, and the number of sequencing reads and sequencing saturation across all cells. The remaining metrics are reported for cells passing QC: the median number of reads per cell, median number of UMIs per cell, median number of genes per cell, median fraction of UMIs mapping to mitochondrial genes, fraction of cell barcodes called as empty droplets and fraction of cell barcodes called as doublets. **(b)** Read mapping QCs. The percent of bases in the sequencing reads (y axis) mapping to the genome, transcriptome and intergenic regions (x axis) across the three protocols (colored bars). **(c)** Cell type assignment. UMAP embedding of single cell profiles from each protocol colored by assigned cell type signature. **(d)** Inferred CNA profiles. Chromosomal amplification (red) and deletion (blue) inferred in each chromosomal position (columns) across the single cells (rows) from the NSCLC-C4 (left) and LE (right) protocols. Top: reference cells not expected to contain CNA in this cancer type. Bottom: cells tested for CNA relative to the reference cells. Color bar: assigned cell type signature for each cell. **(e)** Ambient RNA estimates. Estimates of the fraction of RNA in each cell type derived from ambient RNA contamination (y axis), with cell types ordered by their mean number of UMIs/cell (x axis). Red line: global average of contamination fraction; Green line: LOWESS (locally weighted scatterplot smoothing) smoothed estimate of the contamination fraction within each cell type, along with the associated binomial 95% confidence interval (Clopper–Pearson interval). n = 1 sample per protocol and number of cells (k) is indicated in (a).

**Extended Data Fig. 2. Cell type specific QC metrics for scRNA-Seq protocol comparison in a single NSCLC sample.**
Cell type specific QCs for NSCLC14. Distribution (median and first and third quartiles) of the number of reads per cell, number of UMIs per cell, number of genes per cell and fraction of UMIs mapping to mitochondrial genes in each cell (y axes) in each of the three protocols (x axis), for cells passing QC from each cell type (rows). n = 1 sample per protocol. Number of B cells (k) from NSCLC-C4, PDEC and LE, respectively, is: 100, 121, 78; endothelial cells: 0, 920, 1,078; epithelial cells: 1,284, 641, 260; fibroblasts: 0, 1,476, 1,403; macrophages: 3,306, 911, 727; mast cells: 0, 119, 77; T cells: 449, 723, 722.

**Extended Data Fig. 3. scRNA-Seq protocol comparison for NSCLC following read down-sampling.**
Shown are analyses for NSCLC14 (as in Extended Data Figs. 1 and 2), but after the total number of sequencing reads within each sample was down-sampled to match the protocol with the fewest total sequencing reads. **(a)** Sample processing and QC overview. For each protocol, shown are the number of cells passing QC. The remaining metrics are reported for those cells passing QC: median number of UMIs per cell, median number of genes per cell, median fraction of UMIs mapping to mitochondrial genes in each cell, fraction of cell barcodes called as empty droplets and fraction of cell barcodes called as doublets. **(b, c)** Overall and cell types specific QCs. Distribution (median and first and third quartiles) of the number of UMIs per cell, number of genes per cell and fraction of gene expression per cell from mitochondrial genes (y axes) in each of the three protocols (x axis), for all cells passing QC (b) and for cells from each cell type (c, rows). **(d,e)** Relation of empty droplets and doublets to cell types. UMAP embedding and fraction (horizontal bar) of single cell (gray), empty droplet (red, d) and doublet (red, e) profiles for each protocol **(f)** Cell type assignment. UMAP embedding of single cell profiles from each protocol colored by assigned cell type signature. n = 1 sample per protocol and number of cells (k) is indicated in (a). Number of B cells (k) from NSCLC-C4, PDEC and LE, respectively, is: 114, 157, 78; endothelial cells: 0, 879, 1,078; epithelial cells: 1,283, 644, 260; fibroblasts: 0, 1,439, 1,403; macrophages: 3,278, 853, 727; mast cells: 0, 106, 77; T cells: 432, 663, 722.

**Extended Data Fig. 4. CD45⁺ depletion protocol enriches for non-immune cells in freshly processed NSCLC and ovarian ascites.**
**(a, b)** QCs. Distribution (median and first and third quartiles) of the number of reads per cell, number of UMIs per cell, number of genes per cell and fraction of gene expression per cell from mitochondrial genes (y axes) for all cells passing QC from NSCLC (a) before and after CD45⁺ cell depletion, and for ovarian ascites (k = 2,998 and 10,716 cells, respectively) or (b) after CD45⁺ cell depletion (2,359 cells) (x axis). n = 1 sample per protocol. **(c)** CD45⁺ cell depletion estimates in ovarian cancer ascites by FACS. Flow-cytometry comparison of single cells isolated without (top) or with (bottom) depletion of CD45⁺ cells. Cells were gated by FSC and SSC (first column), doublets removed using FSC-A and FSC-H (second column), 7-AAD gating of dead cells to identify live cells (third column), the distribution of immune and non-immune cells quantified using a CD45 antibody (fourth column) and the distribution of EPCAM⁺ cells quantified using an EPCAM antibody (fifth column). (Efficient removal of CD45⁺ cells from ovarian cancer ascites was also demonstrated with an independent sample from a different patient (data not shown).) Number of cells without and with depletion, respectively are: 10,000, 10,000 (1^st column), 3,468, 2,256 (2^nd column), 3,467, 2,251 (3^rd column), 2,936, 2,174 (4^th and 5^th columns).

**Extended Data Fig. 5. snRNA-Seq protocol comparison in a single neuroblastoma sample.**
(a) Sample processing and QC overview. For each protocol, shown are the number of nuclei passing QC, number of sequencing reads and sequencing saturation across all nuclei. The remaining metrics are reported for those nuclei passing QC: median number of reads per nucleus, median number of UMIs per nucleus, median number of genes per nucleus, median fraction of UMIs mapping to mitochondrial genes in each nucleus and fraction of nucleus barcodes called as doublets. **(b)** Read mapping QCs. The percent of bases in the sequencing reads (y axis) mapping to the genome, transcriptome and intergenic regions (x axis) across the four protocols (colored bars). **(c)** Cell type assignment. UMAP embedding of single nucleus profiles from each protocol colored by assigned cell type signature. **(d)** Inferred CNA profiles. Chromosomal amplification (red) and deletion (blue) inferred in each chromosomal position (columns) across the single nuclei (rows). Top: reference nuclei not expected to contain CNA in this cancer type. Bottom: nuclei tested for CNA relative to the reference nuclei. Color bar: assigned cell type signature for each nucleus. n = 1 sample per protocol and number of nuclei (k) is indicated in (a).

**Extended Data Fig. 6. Cell type specific QC metrics for snRNA-Seq protocol comparison in a single neuroblastoma sample.**
Cell type specific QCs for HTAPP-244-SMP-451. Distribution (median and first and third quartiles) of the number of reads per nucleus, number of UMIs per nucleus, number of genes per nucleus and fraction of UMIs mapping to mitochondrial genes in each nucleus (y axes) in each of the four protocols (x axis), for nuclei passing QC from each cell type (rows). n = 1 sample per protocol. Number of endothelial nuclei (k) from EZ, CST, NST and TST, respectively, is: 69, 32, 91, 95; erythrocyte: 0, 18, 0, 15; T cell: 157, 171, 229, 337; neuroendocrine: 7,379, 5,728, 6,790, 6,477; neural crest: 18, 27, 50, 67; macrophage: 119, 107, 189, 230; fibroblast: 138, 74, 182, 194; zona glomerulosa: 16, 0, 0, 0.

**Extended Data Fig. 7. snRNA-Seq protocol comparison in a single neuroblastoma sample following read down-sampling.**
Shown are analyses for NB HTAPP-244-SMP-451 (as in Extended Data Figs. 5 and 6), but after the total number of sequencing reads within each sample was down-sampled to match the protocol with the fewest total sequencing reads. **(a)** Sample processing and QC overview. For each protocol, shown are the number of nuclei passing QC. The remaining metrics are reported for those nuclei passing QC: median number of UMIs per nucleus, median number of genes per nucleus, median fraction of UMIs mapping to mitochondrial genes in each nucleus and fraction of nucleus barcodes called as doublets. **(b,c)** Overall and cell types specific QCs. Distribution (median and first and third quartiles) of the number of UMIs per nucleus, number of genes per nucleus and fraction of UMIs mapping to mitochondrial genes in each nucleus (y axes) in each of the four protocols (x axis), for all nuclei passing QC (b) and for nuclei from each cell type (c, rows). **(d)** Relation of doublets to cell types. UMAP embedding and fraction (horizontal bar) of single nucleus (gray) and doublet (red) profiles for each protocol. **(e)** Cell type assignment. UMAP embedding of single nucleus profiles from each protocol colored by assigned cell type signature. n = 1 sample per protocol and number of nuclei (k) is indicated in (a). Number of endothelial nuclei (k) from EZ, CST, NST, TST is: 53, 31, 95, 98; erythrocyte: 0, 0, 0, 23; T cell: 146, 177, 230, 345; neuroendocrine: 7,407, 5,726, 6,776, 6,454; neural crest: 14, 27, 50, 69; macrophage: 123, 104, 196, 240; fibroblast: 111, 81, 177, 186; zona glomerulosa: 21, 0, 0, 0.

**Extended Data Fig. 8. Inferred CNA profiles from snRNA-Seq in diverse tumors.**
Chromosomal amplification (red) and deletion (blue) inferred in each chromosomal position (columns) across the single nuclei (rows) from three MBC samples (**a–f**), one ovarian cancer sample (**g-i**) and one sarcoma sample (**j, k**). Top: reference nuclei not expected to contain CNA in this cancer type. Bottom: nuclei tested for CNA relative to the reference nuclei. Color bar: assigned cell type signature for each nucleus. n = 1 sample per protocol and number of nuclei (k) per sample: MBC HTAPP-963-SMP-4741—9,857 (CST), 7,260 (TST); MBC HTAPP-394-SMP-1561—6,948 (CST), 8,058 (NST); MBC HTAPP-589-SMP-2851—7,858 (CST), 8,373 (TST); ovarian HTAPP-316-SMP-991—9,026 (CST), 5,970 (NST), 10,493 (TST); sarcoma HTAPP-951-SMP-4652—7,858 (CST), 4,458 (TST).

**Extended Data Fig. 9. snRNA-Seq protocol comparison of V2 and V3 chemistry from 10x Genomics on a resection of sarcoma.**
**(a)** Sample processing and QC overview. For each protocol, shown are the number of nuclei passing QC, after the total number of sequencing reads from the V3 protocol data was down-sampled to match the number of reads in the V2 data. The remaining metrics are reported for those nuclei passing QC: median number of UMIs per nucleus, median number of genes per nucleus, median fraction of UMIs mapping to mitochondrial genes in each nucleus and fraction of nucleus barcodes called as doublets. **(b)** Overall QCs. Distribution (median and first and third quartiles) of number of UMIs per nucleus, number of genes per nucleus and fraction of UMIs mapping to mitochondrial genes in each nucleus (y axes) for all nuclei passing QC. n = 1 sample per chemistry type and number of nuclei (k) is indicated in (a).

**Extended Data Fig. 10. Comparison of scRNA-Seq and snRNA-Seq from the same tumor sample.**
(**a–g**) CLL. UMAP embedding of scRNA-Seq and snRNA-Seq profiles of the same CLL sample combined by CCA (**Methods**) showing profiles (dots) from both (a), scRNA-Seq (b) and snRNA-Seq (c), colored by assigned cell type signatures. (d) Proportion of cells from each subset in the two protocols. k: number of cells or nuclei passing QC. (e-f) Same UMAP embedding as in (a), colored by cells or nuclei (e) or unsupervised clustering (f). (g) Fraction of cells and nuclei in each cluster. n = 1 sample per protocol and number of cells and nuclei is indicated in (d). (**h–n**) O-PDX neuroblastoma. As in (a-g) for an O-PDX neuroblastoma sample. n = 1 sample per protocol and number of cells and nuclei is indicated in (k). (**o-r**) Dissociation signatures are more prominent in cells than in nuclei from the same tumors. Left and middle: UMAP embedding of scRNA-Seq (left) and snRNA-Seq (middle) profiles (dots) of the same tumor combined by CCA and colored by the score of a dissociation signature (color bar). Right: Distribution of dissociation signature score (y axis; median and first and third quartiles) in cells (green) and nuclei (orange). (o) neuroblastoma (3,449 cells, 7,810 nuclei), (p) MBC (5,163 cells, 7,260 nuclei), (q) CLL (2,562 cells, 2,339 nuclei), (r) O-PDX neuroblastoma (3,495 cells, 4,946 nuclei).

See this image and copyright information in PMC

Cited by

Profiling cell identity and tissue architecture with single-cell and spatial transcriptomics.
Gulati GS, D'Silva JP, Liu Y, Wang L, Newman AM. Gulati GS, et al. Nat Rev Mol Cell Biol. 2025 Jan;26(1):11-31. doi: 10.1038/s41580-024-00768-2. Epub 2024 Aug 21. Nat Rev Mol Cell Biol. 2025. PMID: 39169166 Review.
Integrative insights and clinical applications of single-cell sequencing in cancer immunotherapy.
Liu Z, Li H, Dang Q, Weng S, Duo M, Lv J, Han X. Liu Z, et al. Cell Mol Life Sci. 2022 Oct 31;79(11):577. doi: 10.1007/s00018-022-04608-4. Cell Mol Life Sci. 2022. PMID: 36316529 Free PMC article. Review.
Single-cell and spatial transcriptome assays reveal heterogeneity in gliomas through stress responses and pathway alterations.
He Z, Liu Z, Wang Q, Sima X, Zhao W, He C, Yang W, Chen H, Gong B, Song S, Wang Y. He Z, et al. Front Immunol. 2024 Aug 27;15:1452172. doi: 10.3389/fimmu.2024.1452172. eCollection 2024. Front Immunol. 2024. PMID: 39257581 Free PMC article.
Single-Cell RNA-Seq Reveals Cellular Hierarchies and Impaired Developmental Trajectories in Pediatric Ependymoma.
Gojo J, Englinger B, Jiang L, Hübner JM, Shaw ML, Hack OA, Madlener S, Kirchhofer D, Liu I, Pyrdol J, Hovestadt V, Mazzola E, Mathewson ND, Trissal M, Lötsch D, Dorfer C, Haberler C, Halfmann A, Mayr L, Peyrl A, Geyeregger R, Schwalm B, Mauermann M, Pajtler KW, Milde T, Shore ME, Geduldig JE, Pelton K, Czech T, Ashenberg O, Wucherpfennig KW, Rozenblatt-Rosen O, Alexandrescu S, Ligon KL, Pfister SM, Regev A, Slavc I, Berger W, Suvà ML, Kool M, Filbin MG. Gojo J, et al. Cancer Cell. 2020 Jul 13;38(1):44-59.e9. doi: 10.1016/j.ccell.2020.06.004. Cancer Cell. 2020. PMID: 32663469 Free PMC article.
The Rocky Road from Preclinical Findings to Successful Targeted Therapy in Pleural Mesothelioma.
Paajanen J, Bueno R, De Rienzo A. Paajanen J, et al. Int J Mol Sci. 2022 Nov 3;23(21):13422. doi: 10.3390/ijms232113422. Int J Mol Sci. 2022. PMID: 36362209 Free PMC article. Review.

See all "Cited by" articles

References

1. Cieslik M, Chinnaiyan AM. Cancer transcriptome profiling at the juncture of clinical translation. Nat. Rev. Genet. 2018;19:93–109. doi: 10.1038/nrg.2017.96. - DOI - PubMed
1. Filbin MG, et al. Developmental and oncogenic programs in H3K27M gliomas dissected by single-cell RNA-seq. Science. 2018;360:331–335. doi: 10.1126/science.aao4750. - DOI - PMC - PubMed
1. Jerby-Arnon L, et al. A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade. Cell. 2018;175:984–997.e924. doi: 10.1016/j.cell.2018.09.006. - DOI - PMC - PubMed
1. Puram SV, et al. Single-cell transcriptomic analysis of primary and metastatic tumor ecosystems in head and neck cancer. Cell. 2017;171:1611–1624.e1624. doi: 10.1016/j.cell.2017.10.044. - DOI - PMC - PubMed
1. Tirosh I, et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature. 2016;539:309–313. doi: 10.1038/nature20123. - DOI - PMC - PubMed

Publication types

Actions
Actions
Actions

MeSH terms

Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions

Grants and funding

LinkOut - more resources

Full Text Sources
Other Literature Sources
- The Lens - Patent Citations Database
Medical
- MedlinePlus Health Information
Molecular Biology Databases
- NIAID Data Ecosystem - Find datasets on Infectious and Immune-mediated Diseases
Miscellaneous
- Biospecimen Research Database

[1] Cieslik M, Chinnaiyan AM. Cancer transcriptome profiling at the juncture of clinical translation. Nat. Rev. Genet. 2018;19:93–109. doi: 10.1038/nrg.2017.96. - DOI - PubMed

[2] Cieslik M, Chinnaiyan AM. Cancer transcriptome profiling at the juncture of clinical translation. Nat. Rev. Genet. 2018;19:93–109. doi: 10.1038/nrg.2017.96. - DOI - PubMed

[3] Filbin MG, et al. Developmental and oncogenic programs in H3K27M gliomas dissected by single-cell RNA-seq. Science. 2018;360:331–335. doi: 10.1126/science.aao4750. - DOI - PMC - PubMed

[4] Filbin MG, et al. Developmental and oncogenic programs in H3K27M gliomas dissected by single-cell RNA-seq. Science. 2018;360:331–335. doi: 10.1126/science.aao4750. - DOI - PMC - PubMed

[5] Jerby-Arnon L, et al. A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade. Cell. 2018;175:984–997.e924. doi: 10.1016/j.cell.2018.09.006. - DOI - PMC - PubMed

[6] Jerby-Arnon L, et al. A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade. Cell. 2018;175:984–997.e924. doi: 10.1016/j.cell.2018.09.006. - DOI - PMC - PubMed

[7] Puram SV, et al. Single-cell transcriptomic analysis of primary and metastatic tumor ecosystems in head and neck cancer. Cell. 2017;171:1611–1624.e1624. doi: 10.1016/j.cell.2017.10.044. - DOI - PMC - PubMed

[8] Puram SV, et al. Single-cell transcriptomic analysis of primary and metastatic tumor ecosystems in head and neck cancer. Cell. 2017;171:1611–1624.e1624. doi: 10.1016/j.cell.2017.10.044. - DOI - PMC - PubMed

[9] Tirosh I, et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature. 2016;539:309–313. doi: 10.1038/nature20123. - DOI - PMC - PubMed

[10] Tirosh I, et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature. 2016;539:309–313. doi: 10.1038/nature20123. - DOI - PMC - PubMed

Save citation to file

Email citation

Add to Collections

Add to My Bibliography

Your saved search

Create a file for external citation management software

Your RSS Feed

A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors

Affiliations

A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors

Authors

Affiliations

Erratum in

Abstract

Conflict of interest statement

Figures

Similar articles

Cited by

References

Publication types

MeSH terms

Grants and funding

LinkOut - more resources

Full Text Sources

Other Literature Sources

Medical

Molecular Biology Databases

Miscellaneous

Erratum in

Abstract

Conflict of interest statement

Figures

Similar articles

Cited by

References

Publication types

MeSH terms

Related information

Grants and funding

LinkOut - more resources

Full Text Sources

Other Literature Sources

Medical

Molecular Biology Databases

Miscellaneous