Microscopy, IHC, IF, FISH, RNAscope, and Spatial Assays
TL;DR
Tissue methods answer the question sequencing often loses: where is the signal? IHC detects proteins in tissue. IF uses fluorescent antibody signals. FISH detects DNA/RNA sequences in cells or tissue. RNAscope detects RNA transcripts in situ. Spatial transcriptomics and multiplex imaging extend this into coordinate-aware molecular maps. Sources: [1], [2], [3]
Method comparison
| Method | Detects | Typical use |
|---|---|---|
| H&E | morphology | diagnosis, tumor content, necrosis |
| IHC | protein | lineage markers, PD-L1, ER/PR/HER2 |
| IF | protein with fluorescence | colocalization, multiplex panels |
| FISH | DNA/RNA sequence or locus | amplifications, rearrangements |
| RNAscope | RNA transcript in tissue | gene expression with morphology |
| Spatial transcriptomics | many RNAs with coordinates | tumor microenvironment mapping |
| Multiplex imaging | many proteins/cells in tissue | immune contexture, neighborhoods |
What it does well
- preserve morphology
- assign signals to compartments
- measure heterogeneity
- identify invasive fronts and immune niches
- connect cell state to tissue location
- validate single-cell findings in tissue
What it does not solve
- whole-genome discovery by itself
- quantitative abundance without calibration
- causality
- viability of dissociated cells
- perfect comparability across fixation, staining, scanners, and analysis pipelines
Computational output
| Assay | Data object |
|---|---|
| IHC/IF | image, segmentation masks, intensity scores |
| FISH/RNAscope | spots per cell, colocalization, ROI-level counts |
| Spatial transcriptomics | spot/cell-by-gene matrix + coordinates |
| Multiplex imaging | cell table with marker intensities and neighborhoods |
Developer notes
- Whole-slide images are large; preserve pyramid levels and scanner metadata.
- Segmentation errors can dominate downstream biology.
- Batch effects can come from fixation, staining, antibody lot, scanner, and image processing.
- Spatial coordinates are meaningful only with tissue orientation and ROI context.
- Spot-level spatial transcriptomics may mix multiple cells; cell-level claims require caution.
- Store raw images, masks, feature tables, and processing versions together.
References
- Deng H. Utility of Immunohistochemistry in the Diagnosis of Pleuropulmonary and Mediastinal Cancers: A Review and Update. Arch Pathol Lab Med 2024;148:267-283. PMID 37406295. https://doi.org/10.5858/arpa.2022-0483-RA
- Wang F, Flanagan J, Su N, et al. RNAscope: a novel in situ RNA analysis platform for formalin-fixed paraffin-embedded tissues. J Mol Diagn 2012;14:22-29. PMID 22166544. https://doi.org/10.1016/j.jmoldx.2011.08.002
- Maciejewski K, Czerwinska P. Scoping Review: Methods and Applications of Spatial Transcriptomics in Tumor Research. Cancers (Basel) 2024;16:3100. PMID 39272958. https://doi.org/10.3390/cancers16173100