Abstract:A crucial step prior to analysis of single cell transcriptomic data includes filtering out low-quality (e.g., dying) cells and artifacts (e.g., empty droplets). Choosing appropriate quality control parameters leads to correct cell assignment, so that the data for downstream analysis is high-quality and is representative of all processed cell types. These metrics include high and low cut-offs for mitochondrial content, total number of genes, number of different genes, and ribosomal content per cell. They have been found to vary based on cell and tissue type, as well biological variables such as age and disease, and technical variables such as sequencing technology. The aim of this study was to investigate the variation of these quality control metrics in a data set that included stem, progenitor, and differentiated cell types of the hematopoietic, mesenchymal, and endothelial lineages. Furthermore, we set out to determine whether applying standard cut-offs may differentially affect different cell types. We calculated individual filtering parameters for each cell type using the R package ddqc (Subramanian et al. Genome Biol. 2022). Thus, any biological variability based on mitochondrial, ribosomal, and gene counts was accounted for. Notably, distribution of measurements for each of the above quality control metrics varied by cluster, and certain cell-type means were adjacent to commonly used cut-off values. For instance, we observed higher content of mitochondrial genes in osteoblasts, sinusoidal endothelial cells, a subset of adipocyte-primed mesenchymal cells (MSPC), as well as hematopoietic progenitor cells, pre-B, and B cells. On the other hand, subsets of erythroblasts, neutrophils, as well as arteriolar endothelial cells had the lowest mitochondrial content. In addition, lower total gene counts and lower gene complexity were found in adipocyte-primed MSPCs, neutrophils, and megakaryocyte-erythrocyte progenitors, while higher counts were found in granulocyte-monocyte progenitor cells, megakaryocyte/erythroid-primed multipotent progenitor cells and subsets of erythroblasts. Applying universal cutoff values to this data set would greatly reduce, and some cases, eliminate cell types that have high mitochondrial content or lower gene counts and complexity, while leaving those cell types on the opposite spectrum unaffected. This would also skew proportions of cells in downstream analyses, thus affecting biological interpretation of the results. Overall, our results demonstrate biological variability of the quality control metric distributions in a single cell transcriptomic data set that includes diverse bone marrow niche and hematopoietic cell types. This reveals novel biological information regarding distinct cell types as well as cautions that applying standard filtering paraments on such a data set would eliminate some cell types and bias downstream analysis for others. Optimization of filtering criteria will be important when broadening single cell analysis to multiplexed approaches incorporating transcriptome, genome, epigenome, and proteome-level data.

Low Quality Cells Should Be Removed from Single-Cell RNA-Seq Data Analysis

Precision and Accuracy of Single-Cell/Nuclei RNA Sequencing Data

A Contamination Focused Approach for Optimizing the Single-Cell RNA-seq Experiment

ctQC improves biological inferences from single cell and spatial transcriptomics data

Single Nucleus Total RNA Sequencing of Formalin-Fixed Paraffin-Embedded Gliomas.

Revealing the Prevalence of Suboptimal Cells and Organs in Reference Cell Atlases: An Imperative for Enhanced Quality Control

The need to reassess single-cell RNA sequencing datasets: the importance of biological sample processing

Comprehensive Evaluation of Noise Reduction Methods for Single-Cell RNA Sequencing Data

Enhanced recovery of single-cell RNA-sequencing reads for missing gene expression data

A practical handbook on single-cell RNA sequencing data quality control and downstream analysis

Research on the Data Quality Factors Affecting Single-Cell Type Analysis

Goals and Approaches for Each Processing Step for Single-Cell RNA Sequencing Data.

Biology-inspired data-driven quality control for scientific discovery in single-cell transcriptomics

Analysis of Single-Cell RNA-Sequencing Data: A Step-by-Step Guide

Bone Marrow Microenvironment and Hematopoietic Cell Types Are Differentially Affected By Transcriptomic Quality Control Filtering Metrics in Single Cell Data

Tools for the Analysis of High-Dimensional Single-Cell RNA Sequencing Data

Developing an effective quality evaluation strategy of next-generation sequencing for accurate detecting non-small cell lung cancer samples with variable characteristics: a real-world clinical practice

Quality assessment and control of tissue specific RNA-seq libraries of Drosophila transgenic RNAi models

Single-Cell RNA-Seq Technologies and Related Computational Data Analysis.

scRNABatchQC: multi-samples quality control for single cell RNA-seq data

Optimum processing conditions for single cell RNA sequencing on frozen human PBMCs