Tutorial for Spatial Transcriptomics Data Analysis

Organizers:

Shinn-Ying Ho, National Yang Ming Chiao Tung University, Taiwan
Tzong-Yi Lee, National Yang Ming Chiao Tung University, Taiwan

Speakers:

Tzong-Yi Lee, National Yang Ming Chiao Tung University, Taiwan
Chia-Ru Chung, National Central University, Taiwan
Yuxuan Pang, The University of Tokyo
Audience: Researchers and students working on Spatial Transcriptomics


Tutorial Abstract:

In the rapidly evolving field of high-throughput sequencing, spatial transcriptomics has emerged as a transformative technology, offering unprecedented insights into the spatial organization of gene expression within tissues and organisms. This workshop aims to equip participants with the knowledge and tools necessary to harness the power of spatial transcriptomics data and extract meaningful biological insights. The tutorial will be held in interactive manner, after the introduction of principles, methodologies, and challenges associated with spatial transcriptomics data analysis. Through a combination of lectures, hands-on exercises, and interactive discussions, participants will learn how to preprocess spatial transcriptomics datasets, perform quality control assessments, detect spatially co-expressed genes, and visualize spatial gene expression patterns.

Outline and Duration:

From understanding the fundamentals of spatial gene expression profiling to exploring advanced computational techniques, participants will gain a comprehensive understanding of the analytical pipeline required to interpret spatially resolved transcriptomic data effectively. Additionally, we will explore cutting-edge computational approaches for spatial data integration, trajectory inference, and spatial cell type mapping, providing participants with a toolkit to address diverse biological questions across a range of spatial scales and experimental designs. Whether you are a biologist seeking to leverage spatial transcriptomics to uncover the spatial dynamics of gene regulation or a computational scientist interested in developing novel analytical methods for spatial data analysis, this workshop offers a valuable opportunity to enhance your skills and expand your research horizons. Join us as we embark on a journey through the transcriptomic landscape, where spatially resolved insights await discovery.

In this tutorial on spatial transcriptomics, we will focus on the effective processing, analysis, and visualization of spatial transcriptomics data using Seurat, STForte, and Squidpy. Seurat, a comprehensive R package, has been extended to support spatial data with functionalities beyond basic single-cell analysis, including preprocessing, quality control, normalization, and identification of spatially relevant clusters and markers. Meanwhile, STForte and Squidpy, designed to further enhance spatial data analysis, provides tools to integrate spatial information directly into the analysis pipeline, enabling deeper insights into the spatial organization of cells. Both tools employ robust statistical and machine learning methods such as PCA, t-SNE, and UMAP, which are critical for uncovering complex data structures and understanding the spatial context of gene expression. This tutorial aims to provide participants with a comprehensive framework for conducting spatial transcriptomics research. The topics covered in this tutorial are provided as follows:

1. Introduction to Spatial Transcriptomics (ST) (20 mins)
   • Development of Spatial Transcriptomics
   • Challenges and Opportunities in ST data analysis
   • Tools and Databases for ST data analysis
2. Basic ST Data Analysis Pipeline in Seurat (R language) (30 mins)
   • Data Import & Setup: Prepare spatial data for analysis.
   • Quality Control: Remove low-quality spots for accuracy.
   • Normalization: Correct technical variation across spots.
   • Clustering & Visualization: Group regions based on gene expression.
   • Identify Spatial Features: Detect spatially variable genes.
3. Advanced ST Data Analysis in STForte and Squidpy (Python) (40 mins)
   • Spatial region identification and spatial enhancement using STForte :
     • Spatial-aware dimensional reduction
     • Spatial clustering
     • Spatial enhancement through imputation
   • Spatial neighborhood analysis using Squidpy:
     • Centrality scores and Ripley’s statistics
     • Neighborhood enrichment
     • Spatial autocorrelation
   • Spatial cell-cell communication (CCI) using COMMOT
   • Integrating multi-slices spatial transcriptomics using PASTE2 + STForte