Recent advances in spatial biology, multiplex imaging, and artificial intelligence (AI) are transforming drug discovery, diagnosis, and therapeutic interventions. These technologies provide unprecedented insights into the spatial architecture of tissues, revealing the molecular and cellular interactions that drive disease pathogenesis and therapeutic responses. To harness this potential, we developed a spatial intelligence pipeline that integrates these technologies with experimental biology and multimodal clinical data to identify tumor heterogeneity, immune interactions, drug resistance, neuroinflammation, and therapeutic targets. Through case studies in oncology, Alzheimer's disease, and stroke, we demonstrate how spatial intelligence is addressing critical challenges in precision medicine.

Pulmonary fibrosis (PF) is a chronic, progressive disease that represents the final stage of many interstitial lung diseases (ILDs). Using spatial transcriptomic technologies, we investigate the molecular dysregulation at boundaries between regions of advancing fibrosis and structurally normal alveoli. These results provide key insights into the pathobiology of PF.

Regulatory T (Treg) cells are primarily a subset of CD4+ T cells with well-known roles in immune homeostasis. We utilized single-cell RNA-sequencing to reveal Treg heterogeneity and denoted that signature genes of Treg cells in the activation states enriched with transcriptional or chromatin DNA binding factors, suggesting that chromatin remodeling or transcriptional regulation may play a role in activation. We further performed single-nucleus multi-omics analysis and identified enrichment of binding motifs for specific transcription factor (TF) superfamily in Tregs upon stimulation. Lastly, CUT&RUN, gene perturbation, and spatial ATAC-seq confirmed that the members of the identified TF family regulate selective molecular features in activated Treg cells.

This presentation will explore the rapidly evolving field of spatial biology, focusing on its market dynamics, clinical applications, and future trends. Key points include analysis of market growth projections and regional trends; examination of spatial biology's impact on oncology research, neurodegenerative disease studies, and drug discovery; and discussion of future challenges, including the integration of AI and machine learning, standardization efforts, and ethical considerations in molecular profiling.

This presentation will cover how emerging technologies in spatial biology and single-cell analysis promise to enable better understanding of biological pathways, discovery of new biomarkers and drug targets, and more insightful translational research.

Using image-guided cell sorting systems supported by AI, we analyze morphological features and identify morphological biomarkers to relate cell morphologies to their genomic and proteomic characteristics and predict cell fate and detect cell histories, enabling the system to analyze the past, current, and future characteristics of cells. Hardware advances will also be discussed as they are key to generating high quality data for AI learning.

A major gap in spatial proteomics is the lack of verified mouse antibodies for application to mouse models of disease. Here, we present a novel CODEX/PhenoCycler panel of 28 validated antibodies for murine FFPE tissues. We will discuss our workflow and analysis pipeline and present data from studying the impact of reprogramming nanoparticle gene delivery nanoparticles on the murine melanoma TME.

Fibroblasts play a crucial role in driving pathology across inflammatory and fibrotic diseases, as well as various solid cancer indications. Previously, we identified a Pi16 expressing fibroblast subset found across multiple tissues and diseases that functions as a reservoir, giving rise to activated fibroblasts.  Using multimodal single cell datasets from preclinical models, we examine Pi16 fibroblasts' role in driving activated phenotypes, transcriptional regulators involved, and potential therapeutic opportunities.

There are many choices and trade-offs when selecting an in situ spatial transcriptomic panel. We compare data from two platforms with small (400-500) and large (5,000-6,000) gene panels, including targeted genes, across multiple tissues to better understand how sensitivity, custom additions, and gene coverage influence detection under different biological scenarios.

We quantified the proteomes of 5,883 single cells from human testis using distinct mass spectrometry methods and developed BayesPG, a Bayesian model to systematically infer protein-mRNA discordances. BayesPG estimated consensus levels for 3,861 gene products, 28% of which exhibited significant discordances, contributing to ~1500 GO groups across 6 cell types. We observe that specialized, context specific functions, such as those related to spermatogenesis are regulated after transcription.