Sampling time-dependent artifacts in single-cell genomics studies

Massoni-Badosa, Ramón; Iacono, Giovanni; Moutinho, Cátia; Kulis, Marta; Palau, Núria; Marchese, Domenica; Rodríguez-Ubreva, Javier; Ballestar, Esteban; Rodríguez-Esteban, Gustavo; Marsal, Sara; Aymerich, Marta; Colomer, Dolors; Campo, Elı́as; Juliá, Antonio; Martín‐Subero, José I.; Heyn, Holger

doi:10.1186/s13059-020-02032-0

Cited by 67 publications

(68 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In an epigenetic analysis based on scATAC-seq, we analyzed the PBMC from SCPs, MPs, and HCs, taking three cases from each group. The clustering analysis identified 12 distinct clusters composed of T (CD3G), NK (NKG7), B (MS4A1), and monocyte cells (IL1B) by those signature genes ( 30 ) ( Figure 1A ). We further increased the resolution so that T cells were subclustered into CD8 + T cells and CD4 + T cells ( Figure 1A and Supplementary Figures 2A,B ).…”

Section: Resultsmentioning

confidence: 99%

Epigenetic Landscapes of Single-Cell Chromatin Accessibility and Transcriptomic Immune Profiles of T Cells in COVID-19 Patients

Ling

et al. 2021

Front. Immunol.

View full text Add to dashboard Cite

T cells play a critical role in coronavirus diseases. How they do so in COVID-19 may be revealed by analyzing the epigenetic chromatin accessibility of cis- and trans-regulatory elements and creating transcriptomic immune profiles. We performed single-cell assay for transposase-accessible chromatin (scATAC) and single-cell RNA (scRNA) sequencing (seq) on the peripheral blood mononuclear cells (PBMCs) of severely ill/critical patients (SCPs) infected with COVID-19, moderate patients (MPs), and healthy volunteer controls (HCs). About 76,570 and 107,862 single cells were used, respectively, for analyzing the characteristics of chromatin accessibility and transcriptomic immune profiles by the application of scATAC-seq (nine cases) and scRNA-seq (15 cases). The scATAC-seq detected 28,535 different peaks in the three groups; among these peaks, 41.6 and 10.7% were located in the promoter and enhancer regions, respectively. Compared to HCs, among the peak-located genes in the total T cells and its subsets, CD4+ T and CD8+ T cells, from SCPs and MPs were enriched with inflammatory pathways, such as mitogen-activated protein kinase (MAPK) signaling pathway and tumor necrosis factor (TNF) signaling pathway. The motifs of TBX21 were less accessible in the CD4+ T cells of SCPs compared with those in MPs. Furthermore, the scRNA-seq showed that the proportion of T cells, especially the CD4+ T cells, was decreased in SCPs and MPs compared with those in HCs. Transcriptomic results revealed that histone-related genes, and inflammatory genes, such as NFKBIA, S100A9, and PIK3R1, were highly expressed in the total T cells, CD4+ T and CD8+ T cells, both in the cases of SCPs and MPs. In the CD4+ T cells, decreased T helper-1 (Th1) cells were observed in SCPs and MPs. In the CD8+T cells, activation markers, such as CD69 and HLA class II genes (HLA-DRA, HLA-DRB1, and HLA-DRB5), were significantly upregulated in SCPs. An integrated analysis of the data from scATAC-seq and scRNA-seq showed some consistency between the approaches. Cumulatively, we have generated a landscape of chromatin epigenetic status and transcriptomic immune profiles of T cells in patients with COVID-19. This has provided a deeper dissection of the characteristics of the T cells involved at a higher resolution than from previously obtained data merely by the scRNA-seq analysis. Our data led us to suggest that the T-cell inflammatory states accompanied with defective functions in the CD4+ T cells of SCPs may be the key factors for determining the pathogenesis of and recovery from COVID-19.

show abstract

Section: Resultsmentioning

confidence: 99%

Epigenetic Landscapes of Single-Cell Chromatin Accessibility and Transcriptomic Immune Profiles of T Cells in COVID-19 Patients

Ling

et al. 2021

Front. Immunol.

View full text Add to dashboard Cite

show abstract

“…Our results encourage the use of mixed models, such as the two-part hurdle model with a random effect (e.g., as implemented in MAST with RE), as a way to account for repeated observations from an individual while being able to adjust for covariates at the individual level and, if appropriate, at the individual cell level. Additional random effects, such as sampling time, may also be included 35 . Our extensive simulation study provides valuable information for understanding the power of specific designs and can be used in grant reviews as one justification of the design and analyses employed.…”

Section: Discussionmentioning

confidence: 99%

A practical solution to pseudoreplication bias in single-cell studies

2021

View full text Add to dashboard Cite

Cells from the same individual share common genetic and environmental backgrounds and are not statistically independent; therefore, they are subsamples or pseudoreplicates. Thus, single-cell data have a hierarchical structure that many current single-cell methods do not address, leading to biased inference, highly inflated type 1 error rates, and reduced robustness and reproducibility. This includes methods that use a batch effect correction for individual as a means of accounting for within-sample correlation. Here, we document this dependence across a range of cell types and show that pseudo-bulk aggregation methods are conservative and underpowered relative to mixed models. To compute differential expression within a specific cell type across treatment groups, we propose applying generalized linear mixed models with a random effect for individual, to properly account for both zero inflation and the correlation structure among measures from cells within an individual. Finally, we provide power estimates across a range of experimental conditions to assist researchers in designing appropriately powered studies.

show abstract

“…We selected two PBMC datasets from the Ding datasets: the 10X human and the Dropseq human. We used PBMC datasets for this analysis because their cell-to-cell variability has been extensively studied using single-cell technologies as Fluorescence Activated Cell Sorting (FACS) and scRNA-seq [19, 20, 21, 22, 23]. Using these datasets, we measured the proportion of GO terms obtained in the output that were tightly related to the biological system under study.…”

Section: Resultsmentioning

confidence: 99%

Triku: a feature selection method based on nearest neighbors for single-cell data

Ibáñez-Solé

Inza

et al. 2021

Preprint

View full text Add to dashboard Cite

Feature selection is a relevant step in the analysis of single-cell RNA sequencing datasets. Triku is a feature selection method that favours genes defining the main cell populations. It does so by selecting genes expressed by groups of cells that are close in the nearest neighbor graph. Triku efficiently recovers cell populations present in artificial and biological benchmarking datasets, based on mutual information and silhouette coefficient measurements. Additionally, gene sets selected by triku are more likely to be related to relevant Gene Ontology terms, and contain fewer ribosomal and mitochondrial genes. Triku is available at https://gitlab.com/alexmascension/triku.

show abstract

Sampling time-dependent artifacts in single-cell genomics studies

Cited by 67 publications

References 40 publications

Epigenetic Landscapes of Single-Cell Chromatin Accessibility and Transcriptomic Immune Profiles of T Cells in COVID-19 Patients

Epigenetic Landscapes of Single-Cell Chromatin Accessibility and Transcriptomic Immune Profiles of T Cells in COVID-19 Patients

A practical solution to pseudoreplication bias in single-cell studies

Triku: a feature selection method based on nearest neighbors for single-cell data

Contact Info

Product

Resources

About