Halmos Peter, Liu Xinhao, Gold Julian, Chen Feng, Ding Li, Raphael Benjamin J
Department of Computer Science, Princeton University, 35 Olden St., Princeton, NJ 08544, USA.
Center for Statistics and Machine Learning, Princeton University, 26 Prospect Ave., Princeton, NJ 08544, USA.
Cell Syst. 2025 Feb 19;16(2):101160. doi: 10.1016/j.cels.2024.12.001. Epub 2025 Jan 27.
Spatially resolved transcriptomics (SRT) measures mRNA transcripts at thousands of locations within a tissue slice, revealing spatial variations in gene expression and cell types. SRT has been applied to tissue slices from multiple time points during the development of an organism. We introduce developmental spatiotemporal optimal transport (DeST-OT), a method to align spatiotemporal transcriptomics data using optimal transport (OT). DeST-OT uses semi-relaxed OT to model cellular growth, death, and differentiation processes. We also derive a growth distortion metric and a migration metric to quantify the plausibility of spatiotemporal alignments. DeST-OT outperforms existing methods on the alignment of spatiotemporal transcriptomics data from developing mouse kidney and axolotl brain. DeST-OT estimated growth rates also provide insights into the gene expression programs governing the growth and differentiation of cells over space and time.
空间分辨转录组学（SRT）可在组织切片内数千个位置测量mRNA转录本，揭示基因表达和细胞类型的空间变化。SRT已应用于生物体发育过程中多个时间点的组织切片。我们引入了发育时空最优传输（DeST-OT），这是一种使用最优传输（OT）来对齐时空转录组学数据的方法。DeST-OT使用半松弛OT来模拟细胞生长、死亡和分化过程。我们还推导了生长畸变度量和迁移度量，以量化时空对齐的合理性。在对来自发育中的小鼠肾脏和蝾螈大脑的时空转录组学数据进行对齐时，DeST-OT优于现有方法。DeST-OT估计的生长速率还为控制细胞在空间和时间上生长和分化的基因表达程序提供了见解。
