The 'SpaTopic' R package is centered around the 'SpaTopic' algorithm to infer the spatial tissue architectures from multiplexed images.
Details
The package implements a Collapsed Gibbs sampling algorithm to infer topics, corresponding to distinct tissue microenvironments across multiple tissue images. Without obtaining the cell neighborhood info of every single cell, 'SpaTopic' runs much faster than other KNN-based methods on large-scale images.
The main functions in the 'SpaTopic' package
Prepare input
Seurat5obj_to_SpaTopic
Model Inference
SpaTopic_inference
Author
Xiyu Peng pansypeng124@gmail.com