Analytical tools for studying the tumor microenvironment leveraging spatial transcriptomics

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT Understanding the tumor microenvironment (TME) heterogeneity and architecture are key to stratify cancer patients responsive to immunotherapy and clinical outcome. Single-cell RNA sequencing (scRNAseq) is a powerful tool for studying the TME at the single-cell level, however, spatial information between single cells was not preserved in this technology, which is vital in studying the TME. In contrast, use of spatially resolved transcriptomics holds the promise in the understanding of the spatial contexture of the TME because of its power to capture the location of individual cells within the larger tissue architecture. Recently, commercialization of spatial transcriptomic (ST) technologies have allowed researchers to study at an unprecedented level the spatial architecture of the TME. Similar to other “omics” technologies, novel computational tools are urgently needed to decipher and infer biological meaning for these high-dimensional ST data. In addition to the need for methods for analyzing and visualizing ST data in its current form, there is the need to develop methods for the future state of ST, whereby the level of cellular resolution is vastly decreasing to the single cell level. Moreover, the application of multiple assays to one tissue sample is also producing multiple modalities measured on the same sample (e.g., scRNAseq, image-based cytometry methods such as multiplex immunofluorescence) which requires effective methods for data integration. Lastly, many ST studies are completed on multiple samples simultaneous with the goal of correlating TME features with clinical outcome, thus requiring computational tools to unravel the impact of the spatial architecture of the TME on clinical response. In the proposed research, we will tackle these challenges by implementing state-of- the-art statistical and computational methods that account for and leverage the spatial information present in ST data to understanding the TME. We will develop innovative methods for assessing the TME’s composition (Aim 1) and studying co-localization and spatial heterogeneity (Aim 2), along with hardening of the analytical software, spatialGE, for the analysis and visualization of ST data (Aim 3). The statistical and bioinformatics analytical approaches implemented in spatialGE will allow cancer researchers to easily leverage these methods in their studies of the TME. These analytical methods, along with the developed software tools (spatialGE R/Bioconductor package along with web-based software), will be established in collaboration with clinical, translational, and basic science cancer investigators to ensure usability and interpretability of the developed approaches.