In the new release version 25.1 of the Human Protein Atlas, a major dataset based on Deep Visual Proteomics has been integrated in the Single cell Resource to provide mass-spectrometry based proteomics at an unprecedented resolution for a large number of cell types and tissues.

Deep Visual Proteomics (DVP) combines AI-guided cell segmentation, laser microdissection and ultra-high-sensitivity mass spectrometry (MS) to generate cell type resolved protein profiles. These profiles have been integrated with single-cell and single-nuclei RNA sequencing data from HPA to enable systematic RNA-protein comparison across matching cell populations in pan-organ scale.

27 cell types across 14 tissues from a single healthy female donor was analysed using DVP, resulting in detection of almost 14000 proteins across a dynamic range of 8.5 orders of magnitude.

In the Human Protein Atlas the cell type resolved protein data provides a richer comparative view of RNA and protein. In many cases it confirms protein presence and cell-type specificity, and it can also aid in understanding the complex and context-dependent relationship between transcription and protein expression for proteins with lower RNA-protein correlation. The MS-based data can also be used to validate the distribution and expression of proteins that have been difficult to study using antibodies.

Read the bioRxiv article

Link to press release