Most researchers take microscopy images and then describe what they see in adjectives: the cells look bigger, they seem more rounded after treatment. See the analysis in action: https://youtube.com/watch?v=3_ZQGB8qcyY
Most researchers take microscopy images and then describe what they see in adjectives: the cells look bigger, they seem more rounded after treatment.
These are real observations. Morphology is one of the earliest and most sensitive indicators of drug effect, toxicity, and differentiation. But adjectives don't go into a statistical test.
Cytely extracts area, circularity, elongation, perimeter, and multi-channel intensity for every cell. It turns visual intuition into numbers, and numbers into gateable populations. The cells that "look different" become a defined cluster on a scatter plot, one you can quantify, compare across conditions, and reproduce.
Morphology is the cheapest, most information-dense readout available in phenotypic screening. It requires no special reagents beyond a nuclear stain and a cell body marker. AI-driven phenomics is one of the fastest-growing areas in drug discovery, and it starts with reliable per-cell morphology extraction.
How the assay runs
Segment nuclei from DAPI → optionally expand to approximate cell boundary or use a second channel for cell body segmentation → measure morphological and intensity features per cell across all channels → explore and gate interactively using 2D scatter plots.
What you provide
- Ch1: DAPI (nuclear stain)
- Ch2-4: cell body stain, cytoskeletal marker, or protein-of-interest channels (IF or fluorescent tags)
What you get
- Area
- Circularity
- Elongation (major/minor axis ratio)
- Perimeter per cell
- Mean/median/min/max intensity per channel
Ready to run
Nuclear-only morphology, whole-cell morphology (with cell body stain), multi-channel intensity profiling (up to 4 channels), brightfield cell segmentation, interactive population gating for cell type classification, cell spreading area / substrate-spreading assays (phalloidin/F-actin), platelet spreading morphometry.
Inquire for support
Texture features (Haralick), machine-learning-based phenotype classifiers, cell shape embeddings.