Cluster and dataset comparison

Inclusion Criteria for Data in the Single Cell Type Resource

The scRNA-seq dataset was retrieved from published studies based on healthy human tissues. We performed meta-analysis of literature on scRNA-seq and searched single cell databases, including the Single Cell Expression Atlas (https://www.ebi.ac.uk/gxa/sc/home), the Human Cell Atlas (https://www.humancellatlas.org), the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/), the Tabula Sapiens (https://tabula-sapiens-portal.ds.czbiohub.org/), the Allen Brain Atlas (https://portal.brain-map.org/) and the European Genome-phenome Archive (https://www.ebi.ac.uk/ega/). To avoid technical bias and to ensure the single cell datasets can best represent the corresponding tissues, we applied the following criteria for data selection: (1) Single cell RNA sequencing was performed on single cell suspension from tissues without pre-enrichment of cell types; (2) Datasets included >3,000 cells and 20 million read counts; (3) Pseudo-bulk gene expression profiles were highly correlated with bulk RNA-seq profiles. In total, datasets from 30 tissue types and human blood were included. The samples, their references, and cluster details are listed here.

Tabula Sapiens

The Tabula sapiens project (Tabula Sapiens Consortium* et al. (2022)), includes nearly 500,000 cells from 24 different tissues and organs. The data is publicly available (https://tabula-sapiens.sf.czbiohub.org/) and included in the CZ CellxGene tool where you can explore separate cell types across tissues, here exemplified by epithelial cells across the different tissue samples.

Currently, 6 tissues represented in the aggregated HPA single cell type data, which is used for cell type classification, are imported from the Tabula Sapiens (lung, prostate, salivary gland, thymus, tongue and vascular). Additionally, another 13 tissues are now included in the HPA Single Cell Type resource, for comparison and validation of cell cluster expression profiles. For the 6 tissues already represented by Tabula Sapiens, the original clustering is added to the gene detail pages for easy comparison of cluster expression overview.

Tabula sapiens clustering with HPA clustering of the same data

In the case of the 6 tissue types (lung, prostate, salivary gland, thymus, tongue and vascular) represented by Tabula Sapiens data in the HPA aggregated cell type expression profile, the addition of Tabula Sapiens' own clustering details enables comparison and verification of the HPA pipeline robustness.

Lung

DNAI2 is a protein enriched in ciliated cells, highly expressed in the cilia cluster of the lung sample.

PDGFRA is a protein enriched in fibroblasts.

Prostate

KLK3 is a protein enriched in prostatic glandular cells.

CNN1 is a protein enriched in smooth muscle cells.

Salivary gland

MACC1 is a protein enriched in mucus glandular cells of the salivary gland.

LPO is a protein enriched in serous glandular cells of the salivary gland.

Thymus

THEMIS is a protein enriched in T-cells.

Tongue

KRT5 is a protein enriched in basal keratinocytes.

Vascular

SELE is a protein enriched in endothelial cells, and specifically detected in the endothelial cell clusters.

Tabula Sapiens comparison with non-Tabula Sapiens

For the tissues represented by non-Tabula Sapiens data, the addition of Tabula Sapiens clustering data enables a dataset for comparison and result validation. Here, we show examples of expression overview in each of the tissues that are represented by a non-Tabula Sapiens dataset and compare the cell type expression profile with the Tabula Sapiens results. The comparison for these tissues is available for each protein-coding gene at the gene detail page.

Adipose tissue

In the HPA cell type aggregated data, cell data representing the adipose tissue is based on data from Hildreth AD et al. (2021).
LIPE is a protein enriched in adipocytes, confirmed by immunohistochemical staining in the Tissue Atlas.

Bone marrow

In the HPA cell type aggregated data, cell data representing the bone marrow is based on data from He S et al. (2020).
HBD is a erythroid enriched protein, detected in the erythroid cluster of both datasets.

Eye

In the HPA cell type aggregated data, cell data representing the eye is based on data from Menon M et al. (2019). These datasets differ a lot, since the HPA integrated data is only retina, while the Tabula Sapiens includes the whole eye as a tissue, therefore more cell clusters than retinal cells. RHO is a protein enriched in the rod photoreceptor cells or retina.

Heart muscle

In the HPA cell type aggregated data, cell data representing heart muscle is based on data from MacParland SA et al. (2018).
MB is a protein enriched in the cardiomyocytes of the heart.

Kidney

In the HPA cell type aggregated data, cell data representing the kidney is based on data from Liao J et al. (2020).
SLC12A1 is a protein enriched in distal tubular cells and collecting ducts, confirmed by immunohistochemical staining in the Tissue Atlas.

Liver

In the HPA cell type aggregated data, cell data representing the liver is based on data from MacParland SA et al. (2018).
HAO1 is protein enriched in hepatocytes of the liver, consistent specificity to the cell clusters independent of the dataset. HNF1B is a cholangiocyte specific protein, and selective nuclear staining of bile ducts in the liver is verified by immunohistochemistry in the Tissue Atlas.

Lymph node

In the HPA cell type aggregated data, cell data representing lymph node is based on data from He S et al. (2020).
MS4A1 is a B-cell specific protein, where the Tabula Sapiens with more cells included offers a better resolution to what type of B-cells are showing expression.

Pancreas

In the HPA cell type aggregated data, cell data representing the pancreas is based on data from Qadir MMF et al. (2020).
CPA1 is a protein enriched in pancreatic exocrine glandular cells,

Skeletal muscle

In the HPA cell type aggregated data, cell data representing skeletal muscle is based on data from De Micheli AJ et al. (2020).
MYH2 is a protein enriched in skeletal myocytes.

Skin

In the HPA cell type aggregated data, cell data representing the skin is based on data from Solé-Boldo L et al. (2020).
KRT10 is a protein enriched in suprabasal keratinocytes.

Small intestine

In the HPA cell type aggregated data, cell data representing the small intestine is based on data from Wang Y et al. (2020).
ALPI is a protein with elevated expression in proximal enterocytes.

Spleen

In the HPA cell type aggregated data, cell data representing the spleen is based on data from He S et al. (2020).
IGHA2 is a protein enriched in plasma cells.