Unlike bulk proteomics, spatial proteomics keeps the tissue's geometry intact, revealing which cells sit next to which — information that is completely lost once a tissue is dissociated into a cell suspension.
signal(x,y) = expression(x,y) * intensity - noise
- Tissue grid resolution — how finely the tissue section is subdivided into resolvable pixels/cells.
- Antibody panel coverage — how many distinct protein markers are imaged simultaneously (CODEX/IMC panels can exceed 40).
- Signal intensity — raw fluorescence or metal-tag signal strength captured per marker.
- Background/noise — non-specific staining and detector noise that must be distinguished from true signal.
Spatial proteomics is what let tumor immunologists discover that it's not just how many T cells infiltrate a tumor, but exactly where relative to tumor cells, that predicts checkpoint-inhibitor response.