Segmented breast cancer Xenium SpatialExperiment for the PACE manuscript

A SpatialExperiment object containing the public 10x Genomics human breast cancer Xenium dataset. The object holds 313 genes by 126,515 cells, with counts and logcounts assays, per-cell morphology and quality-control metrics, spatial coordinates, and an 18-level scClassify annotation comprising 17 cell sub-states plus an unassigned class. The downstream analysis collapses these sub-states into eight modelled cell types: tumour, macrophage, stromal, myoepithelial, endothelial, T cell, B cell and dendritic cell. This is the raw input for the breast cancer analyses of the manuscript "PACE: proximity-associated changes in expression for imaging-based spatial transcriptomics". It is consumed by the analysis notebooks analysis_bc.qmd, nichede_bc.qmd, spatiomark_bc.qmd and spillover_composite_bc.qmd in the accompanying repository (https://github.com/ecool50/PACE_manuscript), which regenerate every model fit and figure from it. The PACE method itself is distributed as an R package at https://github.com/ecool50/PACE. The underlying transcript data are the public 10x Genomics preview dataset described in Janesick et al. (2023), Nature Communications.