A comparison of data integration methods for single-cell RNA sequencing of cancer samples

Richards, Laura M.; Riverin, Mazdak; Mohanraj, Suluxan; Ayyadhury, Shamini; Croucher, Danielle; Díaz-Mejía, J. Javier; Coutinho, Fiona J.; Dirks, Peter B.; Pugh, Trevor J.

doi:10.1101/2021.08.04.453579

Cited by 9 publications

(11 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, while single-cell RNA-Seq comparisons of engineered versus in vivo cells can distinguish hybrid identity from population heterogeneity (Han et al, 2020;Shapiro et al, 2013;Tang et al, 2011), the methods typically used to integrate scRNAseq data from multiple sources (i.e. when comparing CFE products to a previously published single cell atlas) tend to over-merge cells with distinct cell states (Babcock et al 2021) (Richards et al, 2021). These issues make it challenging to fairly compare CFE products across protocols and studies.…”

Section: Introductionmentioning

confidence: 99%

Platform-Agnostic CellNet (PACNet) enables cross-study meta-analysis of cell fate engineering protocols

Velazquez

Peng

et al. 2022

Preprint

View full text Add to dashboard Cite

The optimization of cell fate engineering protocols requires evaluating their fidelity, efficiency, or both. We previously adopted CellNet, a computational tool to quantitatively assess the transcriptional fidelity of engineered cells and tissues as compared to their in vivo counterparts based on bulk RNA-Seq. However, this platform and other similar approaches are sensitive to experimental and analytical aspects of transcriptomics methodologies. This makes it challenging to capitalizing on the expansive, publicly available sets of transcriptomic data that reflect the diversity of cell fate engineering protocols. Here, we present Platform-Agnostic CellNet (PACNet), which extends the functionality of CellNet by enabling the assessment of transcriptional profiles in a platform-agnostic manner, and by enabling the comparison of user-supplied data to panels of engineered cell types from state-of-the-art protocols. To demonstrate the utility of PACNet, we evaluated a range of cell fate engineering protocols for cardiomyocytes and hepatocytes. Through this analysis, we identified the best-performing methods, characterized the extent of intra-protocol and inter-lab variation, and identified common off-target signatures, including a surprising neural and neuroendocrine signature in primary liver-derived organoids. Finally, we made our tool accessible as a user-friendly web application that allows users to upload their own transcriptional profiles and assess their protocols relative to our database of reference engineered samples.

show abstract

Section: Introductionmentioning

confidence: 99%

Platform-Agnostic CellNet (PACNet) enables cross-study meta-analysis of cell fate engineering protocols

Velazquez

Peng

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…To integrate snRNA-seq data across all of our samples, we used reciprocal principal component analysis (RPCA), as implemented in Seurat 51,128 . First, we identified 2000 highly variable features (genes) across all of the samples to use as integration features using the ‘SelectIntegrationFeatures()’ function, which we passed as anchor features (‘anchor.features’) to the ‘PrepSCTIntegration()’ function.…”

Section: Methodsmentioning

confidence: 99%

Cocaine addiction-like behaviors are associated with long-term changes in gene regulation, energy metabolism, and GABAergic inhibition within the amygdala

Zhou

Guglielmo

et al. 2022

Preprint

View full text Add to dashboard Cite

The amygdala contributes to negative emotional states associated with relapse to drug seeking, but the cell type-specific gene regulatory programs that are involved in addiction are unknown. Here we generate an atlas of single nucleus gene expression and chromatin accessibility in the amygdala of outbred rats with low and high cocaine addiction-like behaviors following a prolonged period of abstinence. Between rats with different addiction indexes, there are thousands of cell type-specific differentially expressed genes and these are enriched for molecular pathways including GABAergic synapse in astrocytes, excitatory, and somatostatin neurons. We find that rats with higher addiction severity have excessive GABAergic inhibition in the amygdala, and that hyperpolarizing GABAergic transmission and relapse-like behavior are reversed by pharmacological manipulation of the metabolite methylglyoxal, a GABAA receptor agonist. By analyzing chromatin accessibility, we identify thousands of cell type-specific chromatin sites and transcription factor (TF) motifs where accessibility is associated with addiction-like behaviors, most notably at motifs for pioneer TFs in the FOX, SOX, and helix-loop-helix families.

show abstract

“…Given this difficulty, we sought to determine if different levels of imbalance between epithelial normal and epithelial tumor compartments can influence the accuracy of PDAC tumor tissue integration and subsequent classification of tumor cells. As scRNA-seq data from tumor tissue is often integrated across multiple biopsy sites, patients, and cohorts 45 , the ability to reliably quantify tumor cells is imperative to the biological validity of subsequent downstream analyses. We 16 preprocessed and annotated tumor cells in the PDAC samples through integration, unsupervised clustering, and marker gene-based annotation (Online Methods).…”

Section: Perturbation Analysis In Pdac Samples Reveals Tumor Compartm...mentioning

confidence: 99%

Section: Perturbation Analysis In Pdac Samples Reveals Tumor Compartm...mentioning

confidence: 99%

The differential impacts of dataset imbalance in single-cell data integration

Maan

Zhang

et al. 2022

Preprint

View full text Add to dashboard Cite

Single-cell transcriptomic data measured across distinct samples has led to a surge in computational methods for data integration. Few studies have explicitly examined the common case of cell-type imbalance between datasets to be integrated, and none have characterized its impact on downstream analyses. To address this gap, we developed the Iniquitate pipeline for assessing the stability of single-cell RNA sequencing (scRNA-seq) integration results after perturbing the degree of imbalance between datasets. Through benchmarking 5 state-of-the-art scRNA-seq integration techniques in 1600 perturbed integration scenarios for a multi-sample peripheral blood mononuclear cell (PBMC) cohort, our results indicate that dataset imbalance has significant impacts on downstream analyses and the biological interpretation of integration results. We observed significant variation in clustering, cell-type classification, marker-gene-based annotation, and query-to-reference mapping in imbalanced settings. Two key factors were found to lead to quantitation differences after scRNA-seq integration - the cell-type imbalance within and between samples (relative cell-type support) and the relatedness of cell-types across samples (minimum cell-type center distance). To account for evaluation gaps in imbalanced contexts, we developed novel clustering metrics robust to sample imbalance, including the balanced Adjusted Rand Index (bARI) and balanced Adjusted Mutual Information (bAMI). Our analysis quantifies biologically-relevant effects of dataset imbalance in integration scenarios, and introduces guidelines and novel metrics for integration of disparate datasets. The Iniquitate pipeline and balanced clustering metrics are available at https://github.com/hsmaan/Iniquitate and https://github.com/hsmaan/balanced-clustering, respectively.

show abstract

A comparison of data integration methods for single-cell RNA sequencing of cancer samples

Cited by 9 publications

References 41 publications

Platform-Agnostic CellNet (PACNet) enables cross-study meta-analysis of cell fate engineering protocols

Platform-Agnostic CellNet (PACNet) enables cross-study meta-analysis of cell fate engineering protocols

Cocaine addiction-like behaviors are associated with long-term changes in gene regulation, energy metabolism, and GABAergic inhibition within the amygdala

The differential impacts of dataset imbalance in single-cell data integration

Contact Info

Product

Resources

About