Metabolomics,Unknown,Transcriptomics,Genomics,Proteomics

Dataset Information

CellMixS: quantifying and visualizing batch effects in single cell RNA-seq data

ABSTRACT: A key challenge in single cell RNA-sequencing (scRNA-seq) data analysis are dataset- and batch-specific differences that can obscure the biological signal of interest. While there are various tools and methods to perform data integration and correct for batch effects, their performance can vary between datasets and according to the nature of the bias. Therefore, it is important to understand how batch effects manifest in order to adjust for them in a reliable way. Here, we systematically explore batch effects in scRNA-seq data from a variety of datasets according to magnitude, cell type specificity and complexity. We developed a cell-specific mixing score (\texttt{cms}) that quantifies how well cells from multiple batches are mixed. By considering distance distributions (in a lower dimensional space), the score is able to detect local batch bias and differentiate between unbalanced batches (i.e., when one cell type is more abundant in a batch) and systematic differences between cells of the same cell type. We implemented the \texttt{cms}, as well as related metrics to detect batch effects or measure structure preservation, in the CellMixS R/Bioconductor package. We systematically compare different metrics that have been proposed to quantify batch effects or bias in scRNA-seq data using real datasets with known batch effects and synthetic data that mimic various real data scenarios. While these metrics target the same question and are used interchangeably, we find differences in inter- and intra-dataset scalability, sensitivity and in a metric's ability to handle batch effects with differentially abundant cell types. We find that cell-specific metrics outperform cell type-specific and global metrics and recommend them for both method benchmarks and batch exploration.

INSTRUMENT(S): Illumina HiSeq 3000

ORGANISM(S): Homo sapiens

SUBMITTER: Mark Robinson

PROVIDER: E-MTAB-9916 | biostudies-arrayexpress |

REPOSITORIES: biostudies-arrayexpress

ACCESS DATA

Similar Datasets

Project description:In vertebrates, all cells except circulating blood cells must adhere to support their normal growth and functions. The adherence to extracellular matrix and/or other cells is critical and adherent cells placed in non-adherent conditions either die or form multicellular spheroids. Placing cells in non-adherent conditions has been used to induce differentiation in teratocarcinoma cells and more recently to form organoids . Because of such important consequences induced by cell adhesion on cell growth and function, the transition between adherent and non-adherent states is rather rare. There are however physiological situations, such as blood cells diapedesis, during which cells that circulate into the blood stream must adhere to the endothelial cells and cross the endothelial barrier to reach target tissues. Another example of transition, from an adherent to a non-adherent state, is observed in the metastasic process, where cells detach from the tumor mass and circulate in the blood and lymphatic vasculature prior to reattaching and extravasating to colonize distant organs. The comparative analysis of the only effects of adherence on cellular functions is complicated by the fact that in many study models the acquisition or loss of adherence induces major alterations in cell physiology that would obscure the effects of the adherence itself. For example, P19 teratocarcinoma cells differentiate in suspension spheroids while they do not in adhering conditions. In this context, the comparison between spheroids and adherent cells would not be a comparison between adherent and non-adherent cells, but between differentiated cells adhering between them and undifferentiated cells adhering on plastic. Mouse macrophage cell lines represent one of the rare experimental models that may be suitable to compare the adherent and non-adherent states. Indeed, they grow equally well under adherent and non-adherent conditions and keep their differentiated functions under both conditions. We therefore decided to use this model to analyze the changes between the adherent and the non-adherent state using a broad approach, based on proteomics.

Dataset Information

CellMixS: quantifying and visualizing batch effects in single cell RNA-seq data

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure