Transforming Spatial Omics Data Into Biologically Actionable Insights

Our work is motivated by inflammatory and autoimmune diseases, particularly pediatric gastrointestinal and liver disorders, where spatial tissue architecture plays a defining role in disease progression and therapeutic response.

Lab Projects (2)

Transcript Misassignment Correction for Spatial Transcriptomics

Cell segmentation errors and boundary uncertainty remain a major source of noise in single-cell spatial transcriptomics. MisTIC (Missegmented Transcript Inference Correction) is a variational Bayesian framework that explicitly models transcript misassignment across neighboring cells to correct segmentation-induced artifacts without requiring re-segmentation. The method improves cell type identification, spatial differential expression, and downstream inference of cell–cell communication.

MisTIC is designed to be platform-agnostic and integrates seamlessly into existing Xenium, CosMx, and MERSCOPE processing pipelines. Benchmarking on synthetic and real datasets demonstrates improved robustness of cell state and interaction analyses in dense or heterogeneous tissue environments.

Spatial Multi-Omics of Fibrotic Cell–Cell Communication in Pediatric Primary Sclerosing Cholangitis

Primary sclerosing cholangitis (PSC) is an autoimmune cholestatic liver disease marked by inflammation and periductal fibrosis. Despite clinical urgency, the cellular crosstalk driving fibrotic niche formation in PSC is poorly understood. In this project, we applied MERSCOPE spatial transcriptomics integrated with bulk RNA sequencing and SomaScan proteomics to resolve fibrotic microenvironments in pediatric PSC and dissect the cellular circuits that promote fibrosis.

Spatial niche analysis revealed that macrophages expressing moderate-to-high levels of CD163 co-localize with cholangiocytes, T cells, and collagen-producing hepatic stellate cells (HSCs) within periductal fibrotic regions. Cell–cell communication analysis identified a fibrogenic interaction program involving driver gene expression in macrophages and effector gene expression in. Higher composite driver and effector gene scores correlated with established fibrosis and elevated liver stiffness in validation cohorts. Plasma proteins encoded by driver genes were elevated in autoimmune liver diseases and associated with liver stiffness measures. These findings suggest localized macrophage-HSC crosstalk activates fibrotic gene programs and highlights potential molecular targets for therapeutic intervention.