Table 1 Single-cell RNA sequencing data sets and processing.

Species

Human

Human

Human

Human

Number of subjects

14

3

16

10

Cell count

113,063

10,071

13,281

16,498

Tissue /design

Fetal liver 7–17 PCW

4–6 PCW

Mainly pre-treated tumors, viable tumor areas

Tumor and adjacent liver, primary and relapsed tumor

scRNAseq Kit

10 × 3’ v2

CEL-seq2

10 × 3’ v2

Preprocessing

Cell ranger alignment on GRCh38 (STAR),

Cells filtered for > 200 detected genes and total mtCount < 20%;

Genes filtered for expressed in > 3 cells

Cell ranger alignment on GRCh38 (STAR),

Cells filtered for > 200 detected genes and > 500 UMIs and total mtCount < 20%;

Genes filtered for not mtGenes and not hspGenes

Cell Ranger alignment on GRCh38,

Cells filtered for > 200 detected UMIs and > 200/ < 8000 detected genes and total mtCount < 10%

Normalization

By sequencing depth scale to 10,000 counts (NormalizeData, LogNormalize method), data feature scaling, variable gene detection, PCA, Louvain graph-based clustering with a resolution of 30 with standard parameters (fetal liver only, Seurat)

By sequencing depth scaled to 10 000 counts (NormalizeData, LogNormalize method), Scaling, variable gene detection (most 2000), PCA of 2000 most variable genes, 50 first PCs were used to calculate a UMAP (resolution parameter 1) (Seurat)

Seurat standard normalization pipeline,

Shared-nearest neighbor clustering obtained a final of 53 clusters that were used to calculate a UMAP (Resolution parameter 3)

(Seurat)

Download source

https://developmental.cellatlas.io/fetal-liver

http://neuroblastomacellatlas.org

http://omic.tech/scrna-hcc/

Download format

Anndata .h5ad format containing count matrix and metadata

Anndata .h5ad format count matrix and metadata

Anndata .h5ad format count matrix and metadata

Reference

Popescu et al.⁴² (Nature)

Kildisiute et al.⁴³ (Science Advances)

Lu et al.⁴⁴ (Nature Communications)

Raw data

E-MTAB-7407 (Array Express)

EGAD00001008345

EGAC00001001616