QC metrics from four 10x Genomics single cell gene expression datasets.

This dataset contains a collection of summary and QC metrics from four publicly available 10x Genomics single cell gene expression datasets:

qc_examples

Format

A data frame with 27,597 rows and 15 variables:

sample

sample name, one of four values; GBM (glioblastoma), HL (Hodgkin's lymphoma), MB (mouse brain) and PBMC (peripheral blood mononuclear cells)

cell_barcode

the 16 nucleotide cell barcode e.g. AAACCCAAGGCGATAC-1

umap_1

UMAP coordinates 1, for visualisation

umap_2

UMAP coordinates 2, for visualisation

seurat_clusters

cell clusters identified with the Louvain algorithm using default parameters implemented in the Seurat package

cell_type

a very rough cell type annotation

umi_count

the number of UMIs detected

log10_umi_count

log10 of the number of UMIs detected

percent_mt

percentage of UMIs mapping to mitochondrial genes

empty_drops_log_prob

the log-probability of observing the cell's count vector under the null model - from DropletUtils::emptyDrops output

empty_drops_p_value

the Monte Carlo p-value against the null model - from DropletUtils::emptyDrops output

empty_drops_fdr

FDR values returned from DropletUtils::emptyDrops output

nuclear_fraction_droplet_qc

nuclear fraction metric calculated with DropletQC

nuclear_fraction_velocyto

nuclear fraction metric calculated using the output from velocyto

flag

assigned cell status, taking one of three values; cell, damaged_cell, empty_droplet

Source

https://support.10xgenomics.com/single-cell-gene-expression/datasets/

Details

  1. Human Glioblastoma Multiforme: 3’v3 Whole Transcriptome Analysis

  2. Hodgkin's Lymphoma, Dissociated Tumor: Whole Transcriptome Analysis

  3. 10k Mouse E18 Combined Cortex, Hippocampus and Subventricular Zone Cells, Dual Indexed

  4. PBMCs from a Healthy Donor: Whole Transcriptome Analysis

To be included in the dataset, cells were required to pass DropletUtils::emptyDrops using the default FDR threshold of 1 and have a maximum of 15 mitochondrial gene content. The included variables are as follows: