📝 Note: These analyses were done on Visium data courtesy of 10x Genomics, but the approaches apply to other grid-based spatial technologies, like Visium HD.

#SpatialTranscriptomics #SingleCell #Bioinformatics #Visium #10xGenomics #CancerResearch #SpatialOmics #SpatialBiology

💡 Bottom line:
If you've got spatial data, use the space!
Classic single-cell tools are a great start, but ST's real power comes from its spatial context. Don't waste it.

Want us to cover other platforms or methods?
Drop your thoughts in the comments — let's make lemonade! 🍋

3. Gene-distance correlation

🎯 Goal: Link gene expression and distance to a structure.

We tested ERBB2 expression vs distance to the tumor margin.
🔵 Inside the tumor = negative distance
🔴 Outside = positive

➡️ ERBB2 expression drops as distance increases — just as expected.

2. Neighborhood analysis

🎯 Goal: See if clusters co-occur in space or avoid each other.

✅ Immune cells (cluster 4) are enriched (z-score>0) near tumor cells (cluster 1).

Why does this matter? It can drive downstream analyses like cell-to-cell communication between clusters.

1. Spatial clustering

🎯 Goal: Find regions with distinct gene expression and spatial coherence.

✅ In a HER2+ breast cancer sample, spatial-aware clustering found immune cells (green) around the tumor, just like the pathologist said.

❌ Expression-only clustering missed it.

Here's our take:
We LOVE ST ❤️. But we get the skepticism. Too often, spatial data is treated like single-cell data with coordinates slapped on. That's like buying lemons and forgetting to make lemonade!

👇 Here are 3 spatially-aware analyses to get the most out of your ST data:

How to squeeze your spatial transcriptomics data? 🍋
Some scientists claim spatial transcriptomics (ST) will replace single-cell technologies. Others say it's just a pricier single-cell with a nice visual output. So...who's right?

A thread on #SpatialTranscriptomics ⬇️