Dear all,

I created a new column for hgnc_symbols in the result object of DESeq, but I don't know how to label individual point(s) in MA-plot for a specific gene or genes with hgnc_symbols instead of ensemble gene ids.

Any help will be appreciated.

spark

Or if you fancy a visit to the tidyverse then this is the equivalent:

library(DESeq2)
library(ggplot2)
library(dplyr)
library(ggrepel)
set.seed(25)
dds <- makeExampleDESeqDataSet()
dds <- DESeq(dds)
res <- results(dds, tidy=TRUE)
res$Symbol <- paste0("Symbol", 1:1000)

res_df <- res %>%
    as.data.frame() %>%
    dplyr::mutate(significant = padj < 0.5)

res_labelled <- res_df %>%
    dplyr::filter(significant)

ggplot(res_df, aes(x=log2(baseMean), y=log2FoldChange, colour=significant)) +
    geom_point() +
    ggrepel::geom_label_repel(data=res_labelled, aes(label=Symbol)) +
    theme_bw()

The code for the MA plot was adapted from: https://www.bioconductor.org/help/course-materials/2015/BioC2015/bioc2015rnaseq.html

Assuming your results are assigned to res, your gene of interest is "Gene", and the column of the data frame mapped to the hgnc_symbols is labelled as symbol:

plotMA(res, ylim = c(-5,5))
with(subset(res, symbol == "Gene"), {
  points(baseMean, log2FoldChange, col = "dodgerblue", cex = 2, lwd = 2)
  text(baseMean, log2FoldChange, "Gene", pos = 2, col = "dodgerblue")
})

For multiple genes you can do:

plotMA(res, ylim = c(-5,5))
with(subset(res, symbol %in% c("Gene1", "Gene2")), {
  points(baseMean, log2FoldChange, col = "dodgerblue", cex = 2, lwd = 2)
  text(baseMean, log2FoldChange, c("Gene1", "Gene2"), pos = 2, col = "dodgerblue")
})