|
Abstract:
|
Recent technology breakthroughs in spatial molecular profiling (SMP), such as spatial transcriptomics sequencing, enables the comprehensive molecular characterization of single cells while preserving spatial and morphological information. In terms of identifying spatially variable (SV) genes, most of the current work builds upon the geostatistical model with Gaussian process that relies on the selection of ad hoc spatial expression kernels. To overcome this challenge and capture more spatial patterns, we introduced a Bayesian modeling framework. Our model dichotomized the complex sequencing count data into latent binary gene expression levels. Then, pattern quantification is considered as spatial correlation estimation via a modified Ising model. We used auxiliary variable MCMC algorithms to sample from the posterior distribution with an intractable normalizing constant. Simulation showed high accuracy in detecting SV genes compared with alternatives. We also applied our model to real datasets and discovered novel spatial patterns that shed light on the biological mechanisms. This statistical methodology presents new perspectives for characterizing spatial patterns from SMP data.
|
