GLAND: A Global and Local Attention‐Based Framework for Detecting Small and Non‐Continuous Spatial Domains

Welcome to the official documentation for GLAND. This tutorial provides a step-by-step guide to processing spatial transcriptomics data and identifying spatial domains.

Installation

• Installation

Tutorial 1: DLPFC

• Tutorial1: DLPFC Domains Detection with GLAND
  • Environment Configuration
  • Data Loading and Preprocessing
  • Model Training and Spatial Clustering
  • Evaluation and Visualization

Tutorial 2: Spatial Outlier Detection

• Tutorial 2: Spatial Outlier Detection (spLOF)
  • Environment Configuration
  • Data Loading and Preprocessing
  • spLOF Calculate
  • spLOF Analyze

Tutorial 3: Hepatic Lobular

• Tutorial 3: Non-continuous Domains Detection with GLAND
  • Environment Configuration
  • Data Loading and Preprocessing
  • Model Training and Spatial Clustering
  • Evaluation and Visualization

Tutorial 4: Breast Cancer Analysis with GLAND (UMAP)

• Tutorial 4: Breast Cancer Analysis with GLAND (UMAP)
  • Environment Configuration
  • Data Loading and Preprocessing
  • Model Training and Spatial Clustering
  • Evaluation and Visualization
  • UMAP

Tutorial 5: Expression Enhancement

• Tutorial 5: Liver Metastasis Domain Detection with GLAND (Expression Reconstruction)
  • Environment Configuration
  • Data Loading and Preprocessing
  • Model Training and Spatial Clustering
  • Evaluation and Visualization
  • Expression Enhancement

About GLAND

GLAND is designed to extract high-level features from spatial transcriptomics data. It is compatible with various platforms, spanning from sequencing-based methods (10x Visium, Stereo-seq, and ST) to imaging-based technologies (MERFISH, osmFISH, BaristaSeq, and STARmap).