Package index • SpatialEcoTyper

Discovering Spatial Ecotypes

SpatialEcoTyper(): Identify Spatial EcoTypes from Single-cell Spatial Data (A Single Sample)

MultiSpatialEcoTyper(): Integrate Multiple Spatial Transcriptomics Datasets to Identify Conserved Spatial Ecotypes

IntegrateSpatialEcoTyper(): Integrate Multiple Spatial Transcriptomics Datasets to Identify Conserved Spatial Ecotypes

GetSpatialMetacells(): Construct Spatial Metacells from Single-Cell Spatial Data

SNF2(): Enhanced Similarity Network Fusion

ComputeFCs(): Compute Cell-Type-Specific Fold Changes (FCs) for Spatial Clusters

GetPCList(): Generate Principal Component (PC) List for Spatial Neighborhoods

getSN(): Construct Similarity Network

GetSNList(): Construct Cell-Type-Specific Similarity Network

Integrate(): Integrate Spatial Clusters From Multiple Samples Via Similarity Network Fusion

Identifying SE-Specific Cell States

Identifying SE-specific cell states via leave-one-sample-out cross-validation.

LoocvPredict(): Perform leave-one-out cross-validation (LOOCV) for SE prediction

ComputeMetrics(): Assess concordance between true and predicted spatial ecotype labels

NMF Model Development for Spatial Ecotype Recovery and Deconvolution

Developing NMF models for recovering spatial ecotypes from single-cell spatial data, scRNA-seq data, Visium spatial data, or bulk tumor expression data.

CreatePseudobulks(): Create Pseudo-bulk Mixtures

NMFGenerateW(): Train SE Deconvolution Model

NMFGenerateWList(): Train Cell Type-Specific NMF Models for Recovering Spatial EcoTypes

AggregateRecoverModels(): Aggregate recovery models across runs

Recovering Spatial Ecotypes

Recovering spatial ecotypes from single-cell spatial transcriptomics, scRNA-seq, Visium, or bulk tumor expression data.

RecoverSE(): Recovery of SEs Using Pretrained NMF Models

DeconvoluteSE(): Infer SE Abundances Using a Pretrained NMF Model

NMFpredict(): Prediction Using Pretrained NMF Model

Validating Spatial Ecotypes

Validating spatial ecotypes in held-out single-cell or spatial transcriptomics data by testing the co-association or spatial autocorrelation of recovered SE-specific cell states.

Colocalization(): Evaluate Cell State Colocalization

Coassociation(): Compute co-association of cell states across samples

CoassociationTest(): Test Co-association Significance Between Cell States

ComputeNormalizedMoranI(): Compute Normalized Moran's I for Spatial Ecotypes

AverageMarkerExpression(): Compute average expression of gene sets across groups

Clustering Analysis

Clustering via NMF.

nmfClustering(): Robust Clustering via NMF (non-negative matrix factorization)

Visualization

SpatialView(): Visualize Spatial Landscape of Cells / Spots

HeatmapView(): Draw Heatmap

CooccurrenceHeatmapView(): Colocalization Heatmap Visualization

drawRectangleAnnotation(): Draw Rectangular Annotations for Matching Row/Column Groups in a Heatmap

getColors(): Generate a List of Colors

Data Processing

PreprocessST(): Preprocess Spatial Transcriptomics Data

AnnotateCells(): Extract Spatial Ecotype Annotations for Single Cells

Znorm(): Weighted / Unweighted Uni-variance Normalization

rankSparse(): Transform a Sparse Matrix to Rank Space (Rank Non-zeros in Each Column)

matrixMultiply(): Matrix Multiplication with Minibatching and Parallel Processing

fillspots(): Handle Missing Values

mostFrequent(): Identify the most frequent category in a vector

InferNCells(): Infer Cell Numbers per Spot from Expression Profiles