Supplementary Information for Spatial Reconstruction of Single-Cell Gene Expression This supplementary text discusses how spatially diverse landmark genes improve Seurat's mapping. The study evaluates Seurat's sensitivity to the number and type of landmark genes used in spatial reference maps. The resolution of spatial mapping depends on the number of spatially unique combinations of landmark gene expression. The study found that a smaller number of landmark genes spanning a diverse range of expression patterns outperformed a larger number of landmark genes with overlapping or redundant patterns. The study tested different numbers of landmark genes and found that spatial diversity in landmarks provides the greatest benefit to Seurat's mapping. However, increasing the number of landmarks, even if they have overlapping expression patterns, continues to improve Seurat's confidence in the resulting mappings. The study also discusses the variability in published in situ images and the removal of EVL cells, which do not express the same group of genes used to construct the spatial map. The study also presents results on mapping confidence, spatial prediction clustering, and archetype prediction. It also discusses the identification of rare subpopulations and the generation of in situ probes. The study provides a detailed analysis of the spatial reconstruction of single-cell gene expression, including the use of landmark genes, the identification of rare subpopulations, and the generation of in situ probes. The study also provides a detailed description of the Seurat analysis, including the installation and use of the Seurat package, the loading of data, the identification of EVL cells, and the imputation of gene expression data. The study also includes a detailed description of the analysis of in situ patterns and the localization of cellular populations.Supplementary Information for Spatial Reconstruction of Single-Cell Gene Expression This supplementary text discusses how spatially diverse landmark genes improve Seurat's mapping. The study evaluates Seurat's sensitivity to the number and type of landmark genes used in spatial reference maps. The resolution of spatial mapping depends on the number of spatially unique combinations of landmark gene expression. The study found that a smaller number of landmark genes spanning a diverse range of expression patterns outperformed a larger number of landmark genes with overlapping or redundant patterns. The study tested different numbers of landmark genes and found that spatial diversity in landmarks provides the greatest benefit to Seurat's mapping. However, increasing the number of landmarks, even if they have overlapping expression patterns, continues to improve Seurat's confidence in the resulting mappings. The study also discusses the variability in published in situ images and the removal of EVL cells, which do not express the same group of genes used to construct the spatial map. The study also presents results on mapping confidence, spatial prediction clustering, and archetype prediction. It also discusses the identification of rare subpopulations and the generation of in situ probes. The study provides a detailed analysis of the spatial reconstruction of single-cell gene expression, including the use of landmark genes, the identification of rare subpopulations, and the generation of in situ probes. The study also provides a detailed description of the Seurat analysis, including the installation and use of the Seurat package, the loading of data, the identification of EVL cells, and the imputation of gene expression data. The study also includes a detailed description of the analysis of in situ patterns and the localization of cellular populations.