Megas Stathis, Wilbrey-Clark Anna, Maartens Aidan, Teichmann Sarah A, Meyer Kerstin B
Wellcome Sanger Institute, Hinxton, Cambridge, United Kingdom.
Cambridge Stem Cell Institute and Department of Medicine, University of Cambridge, Cambridge, United Kingdom.
Annu Rev Physiol. 2025 Feb;87(1):447-470. doi: 10.1146/annurev-physiol-022724-105144. Epub 2025 Feb 3.
Over the last decade, single-cell genomics has revealed remarkable heterogeneity and plasticity of cell types in the lungs and airways. The challenge now is to understand how these cell types interact in three-dimensional space to perform lung functions, facilitating airflow and gas exchange while simultaneously providing barrier function to avoid infection. An explosion in novel spatially resolved gene expression technologies, coupled with computational tools that harness machine learning and deep learning, now promise to address this challenge. Here, we review the most commonly used spatial analysis workflows, highlighting their advantages and limitations, and outline recent developments in machine learning and artificial intelligence that will augment how we interpret spatial data. Together these technologies have the potential to transform our understanding of the respiratory system in health and disease, and we showcase studies in lung development, COVID-19, lung cancer, and fibrosis where spatially resolved transcriptomics is already providing novel insights.
在过去十年中，单细胞基因组学揭示了肺和气道中细胞类型显著的异质性和可塑性。现在面临的挑战是了解这些细胞类型如何在三维空间中相互作用以执行肺功能，促进气流和气体交换，同时提供屏障功能以避免感染。新型空间分辨基因表达技术的爆炸式增长，再加上利用机器学习和深度学习的计算工具，有望应对这一挑战。在这里，我们回顾了最常用的空间分析工作流程，强调了它们的优点和局限性，并概述了机器学习和人工智能的最新进展，这些进展将增强我们对空间数据的解释方式。这些技术共同有可能改变我们对健康和疾病状态下呼吸系统的理解，我们展示了在肺发育、新冠病毒病、肺癌和肺纤维化方面的研究，其中空间分辨转录组学已经提供了新的见解。
