Beyond benchmarking: an expert-guided consensus approach to spatially aware clustering

作者：Jieran Sun, Kirti Biharie, Peiying Cai, Niklas Müller‐Bötticher, Paul Kießling, Meghan A. Turner, Søren Helweg Dam, Florian Heyl, Sarusan Kathirchelvan, Martin Emons, Samuel Gunz, Sven Olaf Twardziok, Amin El Heliebi, Martin Zacharias, SpaceHack 2.0 participants, Søren Helweg Dam, Fadhl Alakwaa, Shahul Alam, Maria Calleja, Yuzhou Chang, Thomas Chartrand, Nigel Chou, Estella Y. Dong, Michael Fletcher, George Gavriilidis, Alexander Kanitz, Sameesh Kher, Louis Benedikt Kümmerle, Francesca A. Luongo, Qirong Mao, Giorgia Moranzoni, Mar M. Moreno, Anastasiia Okhtienko, Lena K. Perry, Lucie Pfeiferová, Daryna Pikulska, Shyam Prabhakar, Rasool Saghaleyni, Zaira Seferbekova, Vipul Singhal, Divya Sitani, Charlotte Soneson, Sebastian Tiesmeyer, Marco Varrone, Siao-Han Wong, Liya Zaygerman, Teresa Zulueta-Coarasa, Roland Eils, Marcel J. T. Reinders, Raphaël Gottardo, Christoph Kuppe, Brian R. Long, Ahmed Essam Mahfouz, Mark D. Robinson, Naveed Ishaque · 发表于：Nature Methods · 年份：2026 · DOI：10.1038/s41592-026-03194-8 · 被引用次数：4 · 研究领域：Single-cell and spatial transcriptomics、Delphi Technique in Research、Gene expression and cancer classification

Spatial omics technologies have revolutionized the study of tissue architecture and cellular heterogeneity by integrating molecular profiles with spatial localization. In spatially resolved transcriptomics, delineating higher-order anatomical structures is critical for understanding how cellular organization affects tissue and organ function. Since 2020, more than 50 spatially aware clustering (SAC) methods have been developed for this purpose. However, the reliability of current benchmarks is undermined by their narrow focus on Visium and brain tissue datasets, as well as incorrect interpretation of manual annotation as ground truth. Here, we present SACCELERATOR, a community-driven, extensible framework that standardizes data formatting, method integration, and metric evaluation, and is designed to rapidly incorporate new methods and datasets. SACCELERATOR currently includes 22 SAC methods applied to 15 datasets spanning 9 technologies and diverse tissue types. Our analysis revealed substantial limitations in the generalizability and reproducibility of SAC methods across tissues and platforms. We also demonstrate that anatomical labels commonly used as ground truths are often biased, potentially error-prone, and, in some cases, unsuitable for benchmarking efforts. Rather than scoring and comparing methods, we propose a consensus-guided workflow that aggregates clustering results to generate consensus representations. Descriptive spatial metrics highlight areas of high entro...