Inferring somatic copy number aberrations from spatial transcriptomics data to study early cancer development

Understanding the spatial organisation of cells undergoing early transformation and their interactions with the local microenvironment is critical to studying the earliest stages of cancer development. Somatic copy number aberrations (SCNAs) are among the earliest genomic alterations observed in many cancers, but their spatial context remains difficult to study using conventional sequencing approaches. We developed a computational framework to infer SCNAs directly from Xenium spatial transcriptomics data, enabling the identification and spatial mapping of cell populations sharing common SCNA profiles. Applying this approach to ovarian, cervical and breast cancer datasets identified spatially coherent SCNA-defined populations that recapitulated known tumour-associated copy number aberrations and pathological states. In high-grade serous ovarian cancer, we identified an intermediate epithelial population within the fallopian tube exhibiting features of early transformation accompanied by immune remodelling. Together, our pipeline establishes a scalable framework for reconstructing the spatial architecture of SCNA-defined clones, providing new opportunities to identify and characterise candidate early transforming cell populations in situ.