---title: "Spatially Variable Genes - Answer Key"author: - Noor Sohaildate: "2026-04-10"license: "CC-BY-4.0"editor_options: markdown: wrap: 72---```{r}#| label: load_libraries_data#| echo: false# Load libraries and datalibrary(Seurat)library(tidyverse)library(qs2)# Increases the size of the default vector (8GB)options(future.globals.maxSize =8000*1024^2)# Set parallelization (multithreading) to speed up calculationslibrary(future)plan(multisession, workers = parallel::detectCores() -1)# Load datasetseurat_rctd <-qs_read("intermediate/12_seurat_RCTD.qs")```# Exercise 11. Pick a different cell type and run `FindSpatiallyVariableFeatures()`. Name the population you chose and print the top 6 genes in the results.```{r}#| label: morans_i_svgs#| fig-width: 15#| fig-height: 6# Vector of unique cell typescell_types <- seurat_rctd$first_type %>%unique() %>%sort()for (ct in cell_types) {# Subset to cell type P5CRC bins crc_sub <-subset(seurat_rctd,subset = (orig.ident =="P5CRC") & (first_type == ct))# Re-calculate HVGs on subset data crc_sub <-FindVariableFeatures(crc_sub)# Scale log-normalized counts crc_sub <-ScaleData(crc_sub)# Run Moran's I on the top 50 highly variable genes crc_sub <-FindSpatiallyVariableFeatures(crc_sub,selection.method ="moransi",verbose =TRUE,features =VariableFeatures(crc_sub)[1:50],image ="P5CRC.008um")# Get top spatially variable features top_svg <-SpatiallyVariableFeatures(crc_sub, method ="moransi") top_svg <- top_svg[1:6] p <-SpatialFeaturePlot(crc_sub,features = top_svg,pt.size.factor =15,image.alpha =0.1, ncol =3)print(ct)print(top_svg)print(p)}```***[Back to Lesson >>](14_svgs.qmd)[Back to Schedule](../schedule/schedule.qmd)
