4_Data visualization

Enhanced seurat plotting

scCustomize

scPubR

Methods of single cell data visualization

Heatmaps

Heatmaps are widely used to present matrix based data. Seurat and monocle do not provide simple methods to create heat maps, all though they are frequently used in publications. Below are tutorials and example code for single cell and pseudobulk heat maps.

Tutorials

Heatmap basics

Single Cell Heatmap

Pseduobulk heatmap

seu_avg<-AverageExpression(seurat_object, assay ="RNA", group.by = "seurat_clusters")

#subset by genes you want to look at

genes<-c("Foxp3", "Tcf7", "Tbx21")

mat <- seu_avg[[1]][genes,] %>% as.matrix()
mat<- t(scale(t(mat)))

h_cols <-rev(brewer.pal(name = "RdYlBu", n = 7))

quantile(mat, c(0.1, 0.2, 0.4, 0.5, 0.6, 0.8, 0.9))
col_fun = circlize::colorRamp2(quantile(mat, c(0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9)), h_cols)

Heatmap(mat, name = "RNA",  col =col_fun, row_names_gp = gpar(fontsize = 12), column_names_gp = gpar(fontsize = 16), column_names_rot = -90, cluster_columns =T, width = 100, cluster_rows = T, show_column_dend = T, show_row_dend = T)

Bar Plots

Colors

Our package viewmastR provides a helpful function to generate beautiful and distinct colors quickly.

   cluster_cols <- viewmastR::sfc(n=10, scramble = F)

We typically use these color palettes for different assays

   rna_cols <- paletteContinuous(n=8)[c(1:3, 6:8)]
   atac_cols <- paletteContinuous(set = "blueYellow")

   pal<- rev(mako(n=9))
   pal<-pal[1:8]
   pal<-c("#ffffff", pal)
   chromvar_cols<-pal
   
   adt_cols <- paletteContinuous(set = "whiteBlue")
   pseudotime_cols<- viridis::plasma(n=10)
   tcr_cols <- c("gray90", rev(brewer.spectral(n =5)))

   t <-BuenColors::jdb_palette("brewer_spectra")
   multiome_cols <- c(t[2],t[3], "#e5e6e1", t[7],t[9])