R/plot_CoGAPS_pattern_dim.R

In jacobheng/cellwrangler: Supplemental toolkit for single-cell RNAseq analysis

#'Plot CoGAPS pattern on a dimensionality reduction plot
#'
#' @description plot_CoGAPS_pattern_dim() plots Pmeans of a CoGAPS pattern on a dimensonality reduction plot
#'
#' @param CoGAPS_res_set a CoGAPS result set. If set to NULL, a Pmeans matrix may be supplied directly as input
#' for the Pattern_set parameter.
#' @param Pattern_set a string specifying the name of a pattern set (e.g. "nP20") in a CoGAPS_res_set. 
#' Alternatively, if CoGAPS_res_set is set to NULL, a Pmeans matrix may be supplied directly.
#' @param annotation an annotation dataframe. Must contain columns with column names corresponding to values
#' in the group_vector. 
#' @param dim_reduction a dataframe specifying the coordinates of a dimensionality reduction method (e.g. 
#' t-SNE, UMAP etc.). Each column should specify one dimension. Rownames should correspond to the rownames
#' of the annotation object. Colnames will be used as axis labels.
#' @param pattern_num numeric; which pattern to plot.
#' @param facet_wrap_by a string specifying variable for facet_wrap; defaults to NULL.
#' 
#' @keywords plot_pattern_dim
#' @export
#' @return a ggplot object
#' @examples
#' plot_CoGAPS_pattern_dim(CoGAPS_res_set= my_CoGAPS_res, Pattern_set="nP30",  annotation=pData(cds), 
#' dim_reduction = UMAP_proj, pattern_num = 7)



#plot_CoGAPS_pattern_dim function
plot_CoGAPS_pattern_dim <- function(CoGAPS_res_set, Pattern_set, annotation, dim_reduction, pattern_num, 
                             color_scale=c("slategray1","red"), facet_wrap_by=NULL) {
  if(is.null(CoGAPS_res_set)==F) {
    Pmeans <- CoGAPS_res_set[[Pattern_set]]@sampleFactors } else { Pmeans <- Pattern_set }
  
  Pmeans <- Pmeans[unique( as.character(rownames(Pmeans)) ),]
  Pmeans_annot <- merge_by_rownames(annotation, Pmeans, all.x = F, all.y=T)
  #Get dim_reduction names
  dim_reduction_names <- colnames(dim_reduction)
  colnames(dim_reduction) <- c("Dimension1","Dimension2")
  #Merge dim reduction with Pmeans_annot
  Pmeans_annot <- merge_by_rownames(Pmeans_annot , dim_reduction, all.x = T, all.y=F)
  for(i in c( (ncol(annotation)+1):(ncol(Pmeans)+ncol(annotation)) )){
    Pmeans_annot[,i]<-as.numeric(Pmeans_annot[,i])
  }
  
  pattern_to_plot <- ncol(annotation) + pattern_num
  
  p <- ggplot(Pmeans_annot, aes(x=Dimension1,y=Dimension2)) + 
        geom_point(color=Pmeans_annot[,pattern_to_plot]) +
        scale_colour_gradientn(colours = color_scale, values=scales::rescale(c(min(Pmeans_annot[,pattern_to_plot]), max(Pmeans_annot[,pattern_to_plot])))) +
    xlab(dim_reduction_names[1]) + ylab(dim_reduction_names[2]) + 
    guides(fill=guide_legend(title=paste("Pattern ",pattern_num,sep=""))) + theme_classic()
  
  if(is.null(facet_wrap_by) == F) {
    p <- p + facet_wrap(facet_wrap_by)
  } else { p <- p }
  
  return(p)
}