Understanding how cells organize, communicate, and evolve within diseased tissues requires technologies that preserve spatial context while capturing molecular complexity. Although single-cell sequencing has transformed our understanding of cellular heterogeneity, conventional dissociation-based methods disrupt native architecture and eliminate positional information essential for studying microenvironmental structure and cell–cell interactions. To address this gap, this dissertation applies and evaluates DBiT-seq (Deterministic Barcoding in Tissue sequencing), a microfluidics-based spatial multi-omics platform capable of profiling RNA and/or protein expression directly within intact tissue sections.

This work is organized into two major experimental applications. First, I developed and optimized a custom Antibody-Derived Tag (ADT) panel to perform spatial proteomic mapping of human fibrotic lung tissue from idiopathic pulmonary fibrosis (IPF). DBiT-seq spatial proteomics revealed discrete immune and stromal cell populations, including macrophage subtypes, fibroblast subpopulations, and transitional phenotypes positioned at the fibrotic front. The analysis demonstrated spatially structured cellular niches and ECM-associated remodeling patterns, providing evidence of immune–stromal signaling and myofibroblast activation within the fibroblast focus. These observations highlight the utility of DBiT-seq proteomics for resolving clinically relevant disease microenvironments that are difficult to capture with bulk or dissociated single-cell profiling alone.

Second, I applied DBiT-seq spatial transcriptomics to human cerebral organoids co-cultured with glioblastoma (GBM) cells to model early and late tumor invasion. Across the early-stage samples, spatial patterns suggested the onset of extracellular matrix remodeling, stemness programs, and emerging metabolic adaptation. In contrast, late-stage samples displayed structured tumor–brain interfaces, widespread reactive astrocytosis, and a shift toward a hybrid metabolic phenotype concentrated in the surrounding host tissue rather than in the transcriptionally subdued tumor core. Attempts to computationally deconvolve spatial cell states using an external GBM organoid single-cell dataset demonstrated the need for a matched reference atlas and high-quality replicates;​ therefore, future work will include generating paired scRNA-seq datasets from both early- and late-stage GLICO tissues to enable cell-type–resolved spatial modeling, signaling inference, and niche reconstruction.

Together, these studies demonstrate the feasibility and value of DBiT-seq for spatially resolving molecular landscapes in both human disease tissue and organoid-based experimental systems. Beyond producing biological insight into IPF pathology and glioblastoma infiltration, this work identifies analytical considerations, technical bottlenecks, and future methodological pathways needed to advance DBiT-seq toward routine application in spatial systems biology. Ultimately, this dissertation contributes an experimental and analytical framework for applying spatial multi-omics to complex tissue environments and strengthens the foundation for future mechanistic and translational studies in fibrosis, cancer, and regenerative biology.