Modeling fragment counts improves single-cell ATAC-seq analysis

Martens, Laura D.; Fischer, David; Theis, Fabian J.; Gagneur, Julien

doi:10.1101/2022.05.04.490536

Cited by 13 publications

(9 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In scATAC-seq data, the most common normalization strategy is binarization of peaks 136 , 140 , 141 . However, this may also remove biological information and therefore modelling of scATAC counts directly has been suggested 142 . Dimensionality reduction methods based on latent semantic indexing (ArchR 140 and Signac 143 ), latent Dirichlet allocation (cisTopic 141 ) and spectral embedding (snapATAC 136 ) were shown to perform best for downstream clustering and cell annotation 135 .…”

Section: Chromatin Accessibilitymentioning

confidence: 99%

Best practices for single-cell analysis across modalities

et al. 2023

Self Cite

View full text Add to dashboard Cite

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.

show abstract

Section: Chromatin Accessibilitymentioning

confidence: 99%

Best practices for single-cell analysis across modalities

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…We used scVI 43 to perform dimensionality reduction and batch effect removal of the spatial transcriptomics and spatial epigenomics samples (Visium, CosMx, MERFISH, and RNA+ATAC). The transcriptomics data was modeled using a zero-inflated negative binomial distribution 43 , while the chromatin accessibility data was modeled with a Poisson distribution 85 . Spatial proteomics, which quantifies the expression of tens or hundreds of proteins, has lower dimensionality, making dimensionality reduction potentially unnecessary.…”

Section: Dimensionality Reduction and Batch Effect Removalmentioning

confidence: 99%

“…We downloaded the Mouse P22 spatial RNA+ATAC dataset and independently ran dimensionality reduction on each modality using scvi-tools 93 . For ATAC-seq, we selected the fragments expressed in at least 1% of the spots and executed scVI with n_hidden = 334 and n_latent = 18, determined automatically as the square of the number of unique fragments and the square of n_hidden, respectively, n_layers = 1 and employed Poisson likelihood as it was shown to outperform binary approaches 85 . For RNA-seq, we performed CPM and log2 normalization, selected the 5000 most highly variable genes, and ran scVI with n_hidden = 334, n_latent = 18 (to maintain the same embedding size across modalities), and n_layers = 2.…”

Section: Spatial Clustering Of Multi-omics Mouse Brain Datamentioning

confidence: 99%

CellCharter reveals spatial cell niches associated with tissue remodeling and cell plasticity

Varrone

Tavernari

Santamaria-Martínez

et al. 2023

Preprint

View full text Add to dashboard Cite

Tissues are organized in niches where cell types interact to implement specific functions. Spatial -omics technologies allow to decode the molecular features and spatial interactions that determine such niches. However, computational approaches to process and interpret spatial molecular profiles are challenged by the scale and diversity of this data. Here, we present CellCharter, an algorithmic framework for the identification, characterization, and comparison of cellular niches from heterogeneous spatial transcriptomics and proteomics datasets comprising multiple samples. CellCharter outperformed existing methods, identified biologically meaningful cellular niches in different contexts, and discovered spatial cancer cell states, characterized by cell-intrinsic features and spatial interactions between tumor and immune cells. In non-small cell lung cancer, CellCharter revealed a cellular niche composed of neutrophils and tumor cells expressing markers of hypoxia and cell migration. Expression of these markers determined a spatial gradient associated with cancer cell state transition and neutrophil infiltration. Moreover, CellCharter showed that similar compositions of immune cell types can exhibit remarkably different spatial organizations in different tumors, highlighting the need for exploring spatial cell interactions to decipher intratumor heterogeneity. Overall, CellCharter is a flexible and scalable framework to explore and compare the spatial organization of normal and malignant tissues.

show abstract

“…For empirical ATAC-seq data, these regions M are determined by data-dependent peak calling, where peaks are regarded as the set of candidate CREs 27,28 . As snATAC-seq can recover quantitative information on the density and distribution of nucleosomes 24,29 , we use integer values Y cm ∈ {0,1,2, … } to represent the level of accessibility. Existing pipelines diverge in the quantification of snATAC-seq counts, and we propose to use the paired insertion count (PIC) matrix as a uniform input for downstream analyses 24 .…”

Section: Resultsmentioning

confidence: 99%

PACS allows comprehensive dissection of multiple factors governing chromatin accessibility from snATAC-seq data

Miao,

Wang,

Park

et al. 2023

Preprint

View full text Add to dashboard Cite

Single nucleus ATAC-seq (snATAC-seq) experimental designs have become increasingly complex with multiple factors that might affect chromatin accessibility, including cell type, tissue of origin, sample location, batch, etc., whose compound effects are difficult to test by existing methods. In addition, current snATAC-seq data present statistical difficulties due to their sparsity and variations in individual sequence capture. To address these problems, we present a zero-adjusted statistical model, PACS, that can allow complex hypothesis testing of factors that affect accessibility while accounting for sparse and incomplete data. For differential accessibility analysis, PACS controls the false positive rate and achieves on average a 17% to 122% higher power than existing tools. We demonstrate the effectiveness of PACS through several analysis tasks including supervised cell type annotation, compound hypothesis testing, batch effect correction, and spatiotemporal modeling. We apply PACS to several datasets from a variety of tissues and show its ability to reveal previously undiscovered insights in snATAC-seq data.

show abstract

Modeling fragment counts improves single-cell ATAC-seq analysis

Cited by 13 publications

References 32 publications

Best practices for single-cell analysis across modalities

Best practices for single-cell analysis across modalities

CellCharter reveals spatial cell niches associated with tissue remodeling and cell plasticity

PACS allows comprehensive dissection of multiple factors governing chromatin accessibility from snATAC-seq data

Contact Info

Product

Resources

About