Spage2vec is an unsupervised, segmentation-free approach to the spatial transcriptomic diversity of complex tissues, at subcellular resolution. It represents the spatial transcriptomic landscape of a tissue sample as a graph, and uses graph representation learning to create a lower-dimensional representation of local gene expression. These viewers show the spage2vec signatures of four datasets. We thank Mats Nilsson, Sten Linnarsson and Xiaowei Zhuang for making their datasets publicly available.
- In situ sequencing, mouse brain hippocampal area CA1. Data: X. Qian et al., Nature Methods, 2020. doi:10.1038/s41592-019-0631-4
- osmFISH, mouse brain somatosensory cortex. Data: S. Codeluppi et al., Nature Methods, 2018. doi:10.1038/s41592-018-0175-z
- MERFISH, mouse brain hypothalamic preoptic area. Data: J. R. Moffitt et al., Science, 2018. doi:10.1126/science.aau5324
- MERFISH, human fibroblast cells (IMR90). Data: K. H. Chen et al., Science, 2015. doi:10.1126/science.aaa6090
The spatial gene expression data is available on Zenodo, and the code to reproduce the results and figures also on Zenodo.