SpatialROI：在 H&E 图像上自定义感兴趣区域的空间转录组交互分析 R Shiny 平台
SpatialROI: An Interactive R Shiny Platform for Customized Region-Aware Spatial Transcriptomics Analysis
SpatialROI 是一个可在 H&E 图像上直接圈定感兴趣区域（ROI）并完成靶向分析、无需编程的交互式 R Shiny 平台。它支持 10x Visium 的 Space Ranger 原始输出和预处理 Seurat 对象，内置 ROI 绘制、基因集与通路可视化、配体–受体共定位、细胞类型解卷积、spot 聚类、差异表达、样本内分组比较和多样本 ROI 比较共八个分析模块。ROI 可导出为 Seurat 对象或可复用的 spot 索引文件，便于非计算背景用户按组织形态学结构开展区域感知的空间转录组探索。平台免费开放，源代码与示例数据集已公开。
来源：Bioinformatics Advances · doi.org
Motivation
Spatial transcriptomics captures high-throughput molecular information across tissue sections, but histological context remains essential for accurate biological insight. Clustering-derived spatial domains are sensitive to technical noise and may not align with morphologically defined structures that pathologists rely on for disease interpretation. Expert-guided region definition directly on H&E images enables precise identification of molecular signals within tissue architecture. However, existing tools require substantial programming expertise or lack seamless integration between image-based region selection and downstream transcriptomic analysis, limiting accessibility for non-computational users.
Results
We present SpatialROI, an interactive platform that enables users to define regions of interest (ROIs) directly on H&E images and perform targeted analyses without programming. SpatialROI supports standard 10x Visium data as both raw Space Ranger outputs and preprocessed Seurat objects. The platform incorporates eight analytical modules: ROI drawing, gene set and pathway visualization, ligand–receptor colocalization, cell-type deconvolution, spot clustering, differential expression, within-sample group comparison, and multi-sample ROI comparison. ROIs are exportable as Seurat objects or reusable spot-index files. Collectively, SpatialROI facilitates expert-guided, region-aware exploration of spatial transcriptomic data for biologically meaningful insights into tissue architecture.
Availability and implementation
SpatialROI is freely accessible at https://shiny.crc.pitt.edu/spatial_api/. Source code and example datasets are available at https://github.com/myaol/SpatialROI.