Modifying plotMarkerHeatmap to show vertical lines kind of like geom_vline

[munizmom]

Hi ArchR team, I am wondering how can we customize the resulting plot from plotMarkerHeatmap() and draw() to highlight specific regions by adding a vertical line. Seems like adding ggplot nomenclature is not compatible. Do you know how to go around this?
Big thanks in advance and please continue with your great work! Thanks to ArchR we can do an amazing analysis of ATACseq :)
p <- draw(heatmapPeaks_chr19, heatmap_legend_side = "bot", annotation_legend_side = "bot")
p + geom_vline(xintercept = LA_init, linetype="dotdash") + geom_vline(xintercept = LA_end, linetype="dotdash")

Best,
Mar

[rcorces]

plotMarkerHeatmap() uses ComplexHeatmap, not ggplot.
You can check their documentation in case they have the feature you are looking for. Though we dont provide a lot of customization for these heatmaps. In general, if you want something plotted that isnt standard, its either going to require your own custom code to plot or manually editing (e.g. in Adobe Illustrator) afterwards.

[munizmom]

Thanks! I found a roundabout extracting the data from archR project to go to ggplot2 I am giving some info below in case people ask a similar question. In my case I wanted also to plot in the heatmap the info of two clusters and specifically the chr19 coordinates.

library(colorspace); library(ArchR);library(ggplot2);library(dplyr);

#############################################################################################
# How to extract from the output of getMarkerFeatures in ArchR 
#  (stored in the R object called markersPeaks) the info needed to plot the heatmap using ggplot2
# instead of using plotMarkerHeatmap from ArchR
#############################################################################################
markersPeaks <- getMarkerFeatures(
    ArchRProj = atacArch, 
    useMatrix = "PeakMatrix", 
    groupBy = "clusters",
  bias = c("TSSEnrichment", "log10(nFrags)"),
  testMethod = "wilcoxon"
)
#selecting the 2 clusters we want to plot from all the clusters
markersPeaks4astro <- markersPeaks[,c("AF_1","AF_2")]

#Storing the data needed for the creation fo the plot in a dataframe 
log2Df <- as.data.frame(markersPeaks4astro@assays@data$Log2FC)
colnames(log2Df)<-  markersPeaks4astro@colData@rownames

#extracting genomic location info, is going to be the rownames and labels of the plot
Gencoordinates<- as.data.frame(markersPeaks4astro@elementMetadata)
rownames(log2Df) <- paste0(gsub("chr", "", Gencoordinates$seqnames), ": ",
  Gencoordinates$start, "-",Gencoordinates$end);

#selecting chr19 specific region noted as "LA"
log2Df_chr19 <- log2Df[grep("^19: ",rownames(log2Df)),];
#max/min value for the colour scale 
max_Chr19 <- max(log2Df_chr19,abs(min(log2Df_chr19)));


# Trying to annotate beginning end of LA in heatmap
##########################################################
log2Df_chr19$nrow <- 1:length(log2Df_chr19[,1]);
log2Df_chr19$start <- as.numeric(gsub("-.*","",gsub("^19: ","",rownames(log2Df_chr19))));
log2Df_chr19$end <- as.numeric(gsub(".*-","",gsub("^19: ","",rownames(log2Df_chr19))));
# Defining beginning and end positions to add a vertical line in the heatmap
LA_init <- log2Df_chr19[which(log2Df_chr19$start>44000000),"nrow"][1];
LA_end <- tail(log2Df_chr19[which(log2Df_chr19$end<46000000),"nrow"])[6];

#plotting only the peaks in chr19
log2Df_chr19$Label <- rownames(log2Df_chr19);
log2Df_chr19 <- gather(log2Df_chr19, "LA",log2FC,1:2);


# Plotting a heatmap using ggplot 
###########################################################

#colour palette for the heatmap
###############################
mycol2 <-diverging_hcl(length(mycol),
 palette = "Tropic");  
################################

pheatmap <- ggplot(log2Df_chr19, aes(as.factor(Label), LA, fill= log2FC)) + 
  geom_tile() + scale_fill_gradientn(colours = c(mycol2)) + labs(x= "chr19 peaks") +
  theme(axis.text.x = element_text(angle = 45, hjust=1,size=2))+
  geom_vline(xintercept = LA_init, linetype="dotdash") + geom_vline(xintercept = LA_end, linetype="dotdash");

#remove x axis labels
pheatmap2 <- pheatmap + theme(axis.text.x=element_blank(), axis.ticks.x=element_blank());

pdf(file=paste0(wd,f_results,f_plots,"heatmap_Chr19Peaks_v_ggplot.pdf"),onefile=TRUE,width = 8, height =4)
print(pheatmap2);
dev.off();