volcano diagram

[RFFSAH]

Hi, I have a question, I want to creat Volcano-diagram, I want to label the genes and differentiate the upregulated and downregulated by colours? But I have one blue color.

library(GEOquery)
library(limma)
library(umap)

load series and platform data from GEO

tmp <- read.delim("C:/Users/reham/Downloads/GSE18520.top.table.tsv")
View(tmp)

de <- tmp[complete.cases(tmp), ]

ggplot(data=de, aes(x=logFC, y=adj.P.Val)) + geom_point()

p <- ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val))) + geom_point()

p <- ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val))) + geom_point() + theme_minimal()

p2 <- p + geom_vline(xintercept=c(-1, 1), col="red") +
geom_hline(yintercept=-log10(0.01), col="red")

de$diffexpressed <- "NO"
de$diffexpressed[de$log2FC > 1 & de$adj.P.Val < 0.01] <- "UP"
de$diffexpressed[de$log2FC < -1 & de$adj.P.Val < 0.01] <- "DOWN"

p <- ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val), col=diffexpressed)) + geom_point() + theme_minimal()

p2 <- p + geom_vline(xintercept=c(-1, 1), col="red") +
geom_hline(yintercept=-log10(0.01), col="red")

p3 <- p2 + scale_color_manual(values=c("blue", "black", "red"))

mycolors <- c("blue", "red", "black")
names(mycolors) <- c("DOWN", "UP", "NO")
p3 <- p2 + scale_colour_manual(values = mycolors)

de$delabel <- NA
de$delabel[de$diffexpressed != "NO"] <- de$gene_symbol[de$diffexpressed != "NO"]

ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val), col=diffexpressed, label=delabel)) +
geom_point() +
theme_minimal() +
geom_text()

library(ggrepel)
ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val), col=diffexpressed, label=delabel)) +
geom_point() +
theme_minimal() +
geom_text_repel() +
scale_color_manual(values=c("blue", "black", "red")) +
geom_vline(xintercept=c(-1, 1), col="red") +
geom_hline(yintercept=-log10(0.01), col="red")

[AlexisW]

@RFFSAH

ggplot(data=de, aes(x=logFC, y=adj.P.Val)) + geom_point()

@RFFSAH

de$diffexpressed[de$log2FC > 1 & de$adj.P.Val < 0.01] <- "UP"

Is your log-fold-change column named logFC or log2FC? If it's not named log2FC, then these three lines can't work:

de$diffexpressed <- "NO"
de$diffexpressed[de$log2FC > 1 & de$adj.P.Val < 0.01] <- "UP"
de$diffexpressed[de$log2FC < -1 & de$adj.P.Val < 0.01] <- "DOWN"

So in that case the column diffexpressed will have a single value (probably NA as it fails), so that gives a single color.

Make sure your dataframe is correct, then what you had should work:

ggplot(data=de, aes(x=logFC, y=-log10(adj.P.Val), col=diffexpressed)) +
geom_point()