德克萨斯大学吴佳团队宣布他们的研究显示，CoCo-ST在空间转录组学数据集中检测全局和局部生物结构。2025年10月13日出版的《自然—细胞生物学》发表了这项成果。

为了解决这个问题，该课题组人员引入了“比较和对比空间转录组学”（CoCo-ST），这是一种图形对比特征表示框架。通过将目标样本与背景样本进行比较，CoCo-ST可以检测到高方差、广泛共享的结构和低方差、组织特异性的特征。它具有技术优势，包括多样本集成、批处理效果校正以及从现场级Visium数据到单细胞Xenium Prime 5K和亚细胞Visium HD数据的可扩展性。

研究小组将CoCo-ST与十种最先进的空间域检测算法进行了基准测试，这些算法以无主题肺癌前病变样本为主题，证明其识别低方差空间结构的卓越能力被其他方法所忽视。CoCo-ST还能有效区分Visium HD和Xenium Prime 5K数据中的细胞密度和生态位结构。CoCo-ST可在GitHub （https://github.com/WuLabMDA/CoCo-ST）访问。

据了解，空间域检测方法通常关注高变异结构，如肿瘤邻近区域基因表达变化剧烈，而忽略低变异结构，基因表达变化微妙，如邻近正常和早期腺瘤区域之间的结构。

附：英文原文

Title: CoCo-ST detects global and local biological structures in spatial transcriptomics datasets

Author: Aminu, Muhammad, Zhu, Bo, Vokes, Natalie, Chen, Hong, Hong, Lingzhi, Li, Jianrong, Fujimoto, Junya, Chaib, Mehdi, Yang, Yuqiu, Wang, Bo, Poteete, Alissa, Nilsson, Monique B., Le, Xiuning, Cascone, Tina, Jaffray, David, Navin, Nicholas, Wang, Tao, Byers, Lauren A., Gibbons, Don L., Heymach, John, Chen, Ken, Cheng, Chao, Zhang, Jianjun, Wu, Jia

Issue&Volume: 2025-10-13

Abstract: Spatial domain detection methods often focus on high-variance structures, such as tumour-adjacent regions with sharp gene expression changes, while missing low-variance structures with subtle gene expression shifts, like those between adjacent normal and early adenoma regions. Here, to address this, we introduce ‘compare and contrast spatial transcriptomics’ (CoCo-ST), a graph contrastive feature representation framework. By comparing a target sample with a background sample, CoCo-ST detects both high-variance, broadly shared structures and low-variance, tissue-specific features. It offers technical advantages, including multisample integration, batch-effect correction and scalability across technologies from spot-level Visium data to single-cell Xenium Prime 5K and subcellular Visium HD data. We benchmarked CoCo-ST against ten state-of-the-art spatial-domain-detection algorithms using mouse lung precancerous samples, demonstrating its superior ability to identify low-variance spatial structures overlooked by other methods. CoCo-ST also effectively distinguishes cell clusters and niche structures in Visium HD and Xenium Prime 5K data. CoCo-ST is accessible at GitHub (https://github.com/WuLabMDA/CoCo-ST).

DOI: 10.1038/s41556-025-01781-z

Source: https://www.nature.com/articles/s41556-025-01781-z