SpatialEcoTyper: SE discovery from a single sample
MultiSpatialEcoTyper: Integrative analysis of SEs from multiple samples
IntegrateSpatialEcoTyper: Integrative analysis of SEs from multiple samples
RecoverSE: SE recovery
nmfClustering: NMF clustering
SpatialView: draw spatial map of the tissue
HeatmapView: draw a heatmap
CreatePseudobulks: create pseudobulk mixtures
NMFGenerateW: train an NMF model for SE deconvolution from bulk expression profiles
NMFGenerateWList: train cell-type specific NMF model for SE recovery
SpatialEcoTyper 0.0.2
Reduce memory usage by computing distance of spatial neighbors and not all pairwise distances
SpatialEcoTyper 0.0.3
Reduce memory usage by replacing Reduce() with loops
Add seeds to nmfClustering
Re-organize the documentation for integrative analysis
Test and refine all documentations
SpatialEcoTyper 0.0.4
Test Seurat v4.2, v4.4 and v5 for the analysis and add related notes: they lead to different embedding and clustering results, but show high consistency (ARI=0.7) for the demo.
Add more figures to the output directory for integrative analysis
Add hints about the training of SE recovery model: the demo is less robust due to limited number of cells used. The training data should be as comprehensive as possible.
Add minibatch option to SNF2 function to reduce memory usage
Add hints about memory usage and parallel processing
SpatialEcoTyper 0.0.5
Add filter.region.by.celltypes option to SpatialEcoTyper and MultiSpatialEcoTyper
