1282-CODA: Integrative Spatial Transcriptomics Alignment

This article introduces CODA, a computational framework designed to integrate and analyze multi-sample spatial transcriptomics (ST) data. The authors address common technical challenges such as tissue distortion, non-overlapping regions, and batch effects by utilizing a shared low-dimensional feature space for global and local alignment. Unlike existing methods that focus solely on spatial registration, CODA enables downstream spatial analysis to identify spatially consistent genes (SCGs) and spatially differential genes (SDGs). Benchmarking across diverse platforms, including 10x Visium, MERFISH, and Stereo-seq, demonstrates that CODA achieves superior accuracy, computational efficiency, and memory usage. Furthermore, the researchers illustrate CODA's biological utility by reconstructing 3D tissue architectures and uncovering gene expression patterns linked to atherosclerosis and sleep regulation. Ultimately, CODA provides a scalable solution for characterizing spatial heterogeneity in complex biological systems.References:Tan Y, Wang Z, Wang A, et al. Integrative cross-sample alignment and spatially differential gene analysis for spatial transcriptomics[J]. Nature Communications, 2026, 17(1): 5577.前往小宇宙评论区与主播互动