We embarked on this analysis with a stream of millions of RNA transcripts suspended in space. Through rigorous Data Wrangling and Quality Control, we filtered this raw signal to establish a reliable foundation for discovery. In Pre-processing, we navigated the high-dimensional feature space, reducing the data to reveal the underlying biological structure
We then labeled our cells, assigning them to Spatial Domains using novae and defining their Cell Types using a combination of Leiden clustering and HieraType. Just as literal rocket ships are specialized vehicles designed to explore the cosmos, the analysis methods we selected were specialized vehicles for this data. Throughout this guide, we prioritized methods fit for direct hybridization data—such as FOV QC, Novae, and HieraType—rather than applying legacy tools designed for amplification-based technologies.
Yet, our spatial journey has only just begun. So far in this workflow, we have explored Pathway Enrichment within the context of spatial domains. Other tertiary analyses such as ligand-receptor, differential expression, or trajectory inference will be the subject of future chapters in this work, so please check back often.
It is important to highlight that the field of spatial biology is a frontier that is actively expanding. As algorithms improve and new tools emerge, the methods detailed here will undoubtedly evolve. For now, I hope that you have gained a suite of tools and tips to navigate your own discoveries.