Deep Visual Proteomics - Explore the protein profiles in single cell types

NPHS2 - Protein

SMS - Protein

DVP methods

This section of the Single Cell resource contains data generated using Deep Visual Proteomics (DVP). DVP combines high-resolution microscopy with mass spectrometry to enable protein analysis in spatially defined cells within tissue sections. The data was generated by the Mann Lab and collaboratively integrated into the HPA resource. The dataset includes 14 tissue types, represented by 27 distinct cell type-shapes analyzed using DVP. Some tissues are represented by a single cell type, while others include up to four distinct shapes. The complexity of these cell type-definitions varies: some are well-defined, such as hepatocytes and cardiomyocytes, whereas others represent composite structures, such as renal glomeruli. Protein abundance has been quantified across the included cell types, and a specificity classification was performed to identify proteins enriched in particular cell types. The protein data is compared with corresponding single cell RNA data, where matched clusters from the same tissue types and anatomical structures were selected to represent the transcriptomic profiles. Protein detection in each cell type can be explored through interactive bar charts, displayed alongside the selected single cell RNA data, for a matched comparison.

More information about the dataset, including data generation and analysis, is available in the Methods Summary. Details about the included cell type shapes can be found in the DVP details. You can also learn more about the RNA-Protein comparison, see examples and how to further explore using search fields based on enrichment overlap.

Learn about:

Protein detection in single cell types and shapes in human tissues
If a protein is enriched in a particular cell type (specificity)
How RNA and protein overlaps

Bulk tissue data:
The tissues used for DVP, were also used for quantitative bulk MS data generation. Data is avaliable for all 14 tissues, and the result is used to calculate tissue specificity, based on MS data. See more details here.

DEEP VISUAL PROTEOMEⁱ

EXPLORE MOREⁱ Click the links to be directed to a knowledge-summary page with more information and details.
PVALB ALB RNA-Protein comparison DVP details

DVP SUMMARYⁱ

Specificity

Distribution

Classification of all proteins based on single cell type-specific protein detection, determining the number of genes with elevated protein detection (cell type/group enriched, cell type enhanced) in a particular cell type or shape compared to all other cell types and shapes. Click on the bars or names for gene lists with more details.

Cell type group	Cell type	Enriched	Group enriched	Enhanced	Total elevated
Neuronal cells	Neurons	355	1041	718	2114
Glial cells	Astrocytes and Neuropil	68	796	342	1206
	Microglia and Neuropil	109	907	417	1433
Endocrine cells	Pancreatic islets	123	337	297	757
Specialized epithelial cells	Alveolar cell types	126	198	333	657
	Large ducts	123	218	297	638
	Collecting ducts	115	279	202	596
	Hepatocytes	161	238	247	646
	Glomeruli	91	303	222	616
	Proximal tubules	130	349	319	798
	Distal tubules	57	191	190	438
	Granulosa cells	303	459	436	1198
Squamous epithelial cells	Keratinocytes	226	127	185	538
Glandular epithelial cells	Pancreas epithelial cells	58	282	297	637
Ciliated cells	Fallopian tube cells	451	361	267	1079
Germ cells	Oocytes	279	370	302	951
Muscle cells	Cardiomyocytes	21	208	284	513
	Skeletal myofibers	159	153	123	435
Endothelial and mural cells	Capillaries	58	221	150	429
	Smooth muscle cells	175	135	208	518
Blood and immune cells	Neutrophils	184	243	330	757
	Macrophages	83	300	278	661
	T-cells	118	389	458	965
	B-cells	27	255	238	520

DVP - examplesⁱ

Top-enriched in DVP

Protein abundance across different cell types can be used to explore the specificity. Examples of proteins with an enriched detection level across the different DVP samples are shown here: CPS1 show enriched detection in liver hepatocytes, KRT1 in skin keratinocytes. MYBPC1 in skeletal muscle fibers, NPTX2 in pancreatic islets, SLC5A5 in large ducts of salivary gland, TOMM34 in oocytes of ovary and SIGLEC1 in lymph node macrophages. By clicking on the gene-link you will reach the gene summary page where RNA and protein data is displayed side by side for comparison. In addition, immunohistochemical images can be explore in the Tissue resource.

Protein location confirmed by antibody staining, images available in the Tissue resource, confirming the cell and tissue specificity of the DVP-based protein data. Scale bar is 30um

Elevated proteins

Proteomics data is integrated to enable exploration of proteins with cell type enrichment, alongside side-by-side comparison with corresponding RNA expression data. Specificity categories and elevated cell types are searchable, with respective numbers summarized in the tables above. A global overview of cell type enriched proteins is presented in the pie chart, bar plot, and summary table above. Here, we showcase three examples of elevated proteins and how the protein detection profile varies, examples are spatially confirmed by antibody labeling, images available in the Tissue resource.

GZMK - Protein
GZMK - lymph node

GZMK is an example of protein classified as enriched.

BSND - Protein
BSND - salivary gland

BSND is an example of protein classified as group enriched

TJP1 - Protein
TJP1 - kidney

TJP1 is an example of protein classified as enhanced

Low cell type specificity

Detected RNA and protein levels, not classified as elevated (enriched, group enriched or enhanced) are classified as "low cell type specificity". These proteins are generall detected with low variation, usually with a low tau score and often a more general expression profile. Comparing the RNA and protein data across these targets is a challange, and for this reason we integrated the Agreement scores.

Examples of proteins with low cell type specificity

Essential proteins, such as Splicing factor 3b subunit 2 (SF3B2) involved in RNA processing, and Golgin A4 (GOLGA4) are both generally expressed detected in all the DVP samples. Both score 1 in detection agreement, since RNA and protein is detected in all cell types. The expression agreement score is however lower, for both, where GOLGA4 scores 0.41 while SF3B2 scores 0.75.

SF3B2 - Protein
SF3B2 - skin

GOLGA4 - Protein
GOLGA4 - fallopian tube