Title:  AI for Single-cell Spatially Resolved Multiplex Data

Committee: 

Dr. Coskun, Advisor

Dr. Inan, Co-Advisor     

Dr. Dyer, Chair

Dr. Qiu

Abstract: The objective of the proposed research is to use machine learning and deep learning for the analysis and understanding of highly multiplex spatially resolved omics data at the cellular and subcellular level. Spatial omics technologies generate highly multiplexed and multimodal single cell data in their native tissue environment. These technologies require novel analysis methods for understanding single cell heterogeneity within their spatial context. This research target to propose AI-based solutions for analyzing highly multiplex spatially resolved single-cell data. In this research we present (1) cross-modality and within-modality translation of multiplex immunofluorescence imaging at subcellular resolution using conditional generative adversarial neural network, (2) graph-based analysis of cellular and subcellular networks of multiplex imaging in cancer tissue, and (3) multimodal data integration of single-cell protein and metabolomic imaging.