A0442
Title: Spotiphy enables single-cell spatial whole transcriptomics via generative modeling
Authors: Jiyang Yu - St. Jude Children\'s Research Hospital (United States) [presenting]
Abstract: Spatial transcriptomics (ST) has revolutionized the understanding of tissue regionalization by making it possible to visualize gene expression across an intact tissue section, but the approach remains dogged by the challenge of achieving single-cell resolution without sacrificing whole genome coverage. Spotiphy (Spot imager with pseudo-single-cell resolution histology) is presented, a novel computational toolkit that transforms sequencing-based ST data into single-cell-resolved whole-transcriptome images. It achieves this by (i) leveraging generative modeling with single-cell RNA sequencing (scRNA-seq) data and high-resolution histological images to convert spot-level ST data into single-cell whole transcriptomic profiles, (ii) utilizing Gaussian processes to impute the cellular composition and expression profiles of non-capture areas, and then (iii) merging the results of the first two steps into a whole-slide image. In evaluations that used matched scRNA-seq, Visium, Xenium, CosMx, and immunohistochemistry datasets for Alzheimer's Disease and normal mouse brains, Spotiphy delivers the most precise cell-type proportions and accurately depicts the distribution of rare cell populations such as immune cells. By making it possible to visualize the cellular localization and expression profiles of an intact tissue section, Spotiphy will be an important tool for gaining insight into the cellular organization, heterogeneity, and function of complex biological systems.
