Identification of Spatially Variable Genes (SVGs) Open Access

Abstract

Spatial transcriptomics represents a rapidly evolving technology with profound applications in diverse biological and medical fields. The identification of spatially variable genes (SVGs) is crucial due to the vast number of genes, necessitating selection for tissue-specific relevance, as downstream analyses heavily rely on these chosen SVGs.

Here, we introduce a grid-based approach grounded in the Poisson model, providing a simple yet powerful statistical tool applicable to various spatial transcriptomics datasets. Our method demonstrates adaptability across multiple datasets, exhibiting computational efficiency, especially with high-resolution data.

Our methodology accurately identified all 15 SVGs alongside 85 non-SVGs in each pattern through simulations encompassing prevalent gene spatial patterns including gradients and hotspots. Subsequent validation on three real datasets — both image-based and sequencing-based technologies, conventional and high-resolution platforms — further affirmed the efficacy of our approach. We compared our method to SpatialDE, SPARK-X, and SMASH and identified SVGs bearing biological significance.

Overall, our study underscores the utility and adaptability of our method in advancing spatial transcriptomics research, offering valuable insights into spatial gene expression dynamics.

Table of Contents

1 Introduction

2 Background

2.1 Treedsceek

2.2 SpatialDE

2.3 SPARK and SPARK-X

2.4 Others

3 Methods

3.1 Preprocessing and quality control

3.2 Complete Spatial Randomness

3.3 Poisson Model

4 Results

4.1 Simulations

4.2 Application to Human Osteosarcoma

4.3 Application to Mouse Brain

4.4 Application to Human colon

5 Discussion

A Appendix

Bibliography

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	M.S.P.H.
Submission	Master's Thesis
Language	English
Research Field	Health Sciences, General Statistics Biology, Genetics
Keyword	spatially variable gene spatial transcriptomics
Committee Chair / Thesis Advisor	Steve Qin, Emory University
Committee Members	Jian Hu, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Identification of Spatially Variable Genes (SVGs) ()	2024-04-08 01:20:53 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions