发表机构
Sungkyunkwan University; Mass General Brigham, Harvard Medical School; The University of Texas MD Anderson Cancer Center; Dana-Farber Cancer Institute; Broad Institute of Harvard and MIT; Massachusetts Institute of Technology; Cedars-Sinai Medical Center; Harvard University; Samsung Advanced Institute of Health Science and Technology, Sungkyunkwan University（成均馆大学; 麻省总医院布莱根分院，哈佛医学院; 德克萨斯大学MD安德森癌症中心; 丹娜-法伯癌症研究所; 哈佛-麻省理工博德研究所; 麻省理工学院; 西达赛奈医疗中心; 哈佛大学; 三星健康科学技术高级研究院，成均馆大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对虚拟空间转录组学缺乏统一基准的问题，提出STP-BENCH，涵盖六种癌症、两个平台，统一评估21种方法，发现统一形态学编码重排模型排名，并公开基准。
AI 中文摘要
空间转录组学（ST）通过捕获空间分辨的基因表达，为肿瘤异质性提供了前所未有的见解，但其高昂的实验成本阻碍了其大规模应用。因此，直接从苏木精-伊红（H&E）切片预测空间基因表达的计算方法（称为虚拟ST）迅速涌现。尽管取得了进展，但由于基准测试不足，评估该领域的进展仍然困难：先前的研究依赖于小型、异构的数据集，不一致的训练和推理流程，以及对生物学可解释性和模型鲁棒性的评估有限。为解决这些不足，我们提出了STP-BENCH，一个用于虚拟ST模型的标准化基准。STP-BENCH涵盖两个ST平台（Visium和Xenium）上的六种癌症类型，每个训练数据集包含超过30,000个点且至少15张切片，以确保统计可靠性。我们评估了21种预测方法，并在架构适用时，使用统一的病理学基础模型作为形态学编码器重新实现这些方法。除了常规基准中报告的高变基因平均预测准确性外，我们系统地考察了哪些基因和基因集可从组织形态学中恢复。我们进一步通过细胞类型反卷积和空间域识别评估预测谱的下游生物学效用，并评估模型在域偏移和数据缩放下的可靠性。值得注意的是，统一的形态学编码显著重新排序了先前研究中建立的模型排名，表明架构创新和图像编码在之前的评估中被混淆。我们公开发布STP-BENCH以支持可重复性，并作为社区基准，网址为https://这个URL。
英文摘要
Spatial transcriptomics (ST) provides unprecedented insights into tumor heterogeneity by capturing spatially resolved gene expression, yet its high experimental cost hinders large-scale adoption. Consequently, computational approaches that predict spatial gene expression directly from hematoxylin and eosin slides, termed virtual ST, have rapidly emerged. Despite this progress, assessing advances in the field remains difficult due to insufficient benchmarking: prior studies rely on small, heterogeneous datasets, inconsistent training and inference pipelines, and limited evaluation of biological interpretability and model robustness. To address these gaps, we present STP-BENCH, a standardized benchmark for virtual ST models. STP-BENCH comprises six cancer types spanning two ST platforms (Visium and Xenium), with each training dataset containing more than 30,000 spots and at least 15 slides to ensure statistical reliability. We evaluate 21 predictive approaches, re-implemented with a unified pathology foundation model as the morphological encoder when architecturally applicable. Beyond conventional benchmarks that report average predictive accuracy on highly variable genes, we systematically examine which genes and gene sets are recoverable from histomorphology. We further evaluate the downstream biological utility of predicted profiles through cell-type deconvolution and spatial domain identification, and assess model reliability under domain shifts and data scaling. Notably, unified morphological encoding substantially re-orders model rankings established in prior studies, indicating that architectural innovations and image encoding have been conflated in previous evaluations. We publicly release STP-BENCH to support reproducibility and serve as a community benchmark at https://github.com/NEXGEM/STP-Bench.
