A novel statistical learning framework to uncover clinically significant spatial ecotypes with single cell spatial multiomics data

Author

Khoa Le Anh HuynhFollow

DOI

https://doi.org/10.25772/B6CJ-MR64

Defense Date

2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Biostatistics

First Advisor

Jinze Liu

Abstract

Advancements in high-plex spatial omics technologies have revolutionized our ability to map cellular organization and interactions within tissues at near single-cell resolution. However, analyzing and integrating these complex, multi-dimensional datasets requires robust, scalable, and interpretative computational tools. This thesis introduces “A novel statistical learning framework to uncover clinically significant spatial ecotypes with single cell spatial multiomics data”, a comprehensive framework designed to facilitate accurate cell type annotation, spatial neighborhood mapping, intercellular communication analysis, and translational applications. Key innovations include TACIT, an unsupervised machine learning algorithm capable of precise cell type and state deconvolution across diverse tissues; STARComm, a scalable system for identifying spatial co-location of cell-cell communication in both 2D and 3D tissues; and AstroSuite, an integrated platform that combines these tools with advanced visualization and drug-target prediction modules. Applications to human oral mucosa, salivary glands, and cancer tissues demonstrated the pipeline’s capacity to generate detailed tissue atlases, uncover microenvironmental features, and identify potential therapeutic targets. The integrated approach offers opportunities for personalized medicine, enabling tissue-specific biomarker discovery, tailored treatment strategies, and improved prognostic assessments. Overall, this work advances the field of spatial omics by bridging molecular profiling with clinical translational science, fostering the development of precision therapeutics in complex tissue landscapes.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

7-28-2025