We investigated the lateral plate mesoderm (LPM), a bilateral layer which contributes to the development of tissues as diverse as the heart, kidney, and skeletal muscle, as the early organisation of LPM boundaries remains cryptic, with few elements of the gene regulatory network understood. The development programme relies on robust expression of spatially restricted genes which are critical to development, where variation in expression may lead to catastrophic defects and disease, and genome-wide transcript analysis is required to understand these patterns. Recognition of the analytical power generated by these approaches has sparked the emergence of the new field of spatial transcriptomics (ST), however there is a dearth of tools available for effective analysis due to the complexity of ST data.
Using a novel adaptive two-peak statistical model and unsupervised hierarchical clustering techniques, we analysed publicly available Tomo-seq data (Junker et. al 2014) in Danio rerio to unbiasedly identify and characterise unannotated, spatially restricted genes in the LPM domain in 3D. We systematically predicted new genetic elements contributing to gene networks deployed in LPM boundaries, and developed a method to resolve orientational discrepancies at the cryo-sectioning stage, and utilised 3D modelling to validate against in situ imaging data. Our computational approach successfully reconstructed detailed gene modules of expression within the developing LPM, providing new insights into the complex development of embryonic boundaries and novel gene targets for future study. Our study highlights the power of ST, and the promise it holds for uncovering how embryonic spatial boundaries contribute to adult development.