Next generation pipeline for next generation spatial transcriptomic data

Recent years have seen the rapid rise of spatial transcriptomic (ST) technologies. Among them, in situ hybridization-based methods can profile hundreds of genes at subcellular resolution, enabling unprecedented characterization of healthy and diseased tissue architecture. The launch of commercial platforms like Xenium and CosMx has further accelerated the generation of high-quality datasets, whose size and number are growing exponentially.
However, despite these advances, analyzing such datasets remains a major challenge. Current strategies for quality control, normalization, and cell clustering are largely adapted from single-cell genomics, often lacking statistical rigor and rational justification. Moreover, many spatial analysis methods—such as those for detecting spatially variable genes—were designed for small-scale datasets and are not scalable to datasets containing millions of cells.
To address these challenges, we developed TranspaceR, an R package designed for the efficient analysis of large-scale in situ hybridization-based ST datasets. Built on robust statistical principles, TranspaceR enables reliable and scalable analysis, paving the way for new insights into tissue organization.