A Hidden Markov Random Field Model for Detecting Domain Organizations from Spatial Transcriptomic Data
Cells in complex tissues are organized by distinct microenvironments and anatomical structures. This spatial environment of cells is thought to be important for division of labor and other specialized functions of tissues. Recently developed spatial transcriptomic technologies enable the quantification of expression of hundreds of genes while accounting for cells’ spatial coordinates, providing an opportunity to study spatially organized structures. Here, we describe a computational pipeline for detecting the spatial organization of cells based on a hidden Markov random field model. We illustrate this pipeline with data generated from multiplexed smFISH from the adult mouse visual cortex.
Key wordsHidden Markov random field Spatial organization Sequential fluorescence in situ hybridization Multiplexed fluorescence in situ hybridization
- 14.Wang Q (2012) HMRF-EM-image: implementation of the hidden markov random field model and its expectation-maximization algorithm. arXiv PreprGoogle Scholar
- 15.Li SZ (2009) Markov random field modeling in image analysisGoogle Scholar
- 16.Li SZ (2003) Modeling image analysis problems using Markov random fields. Stoch Process Model Simul 473Google Scholar