EnhancedVolcano selectLab issue

Question

[renhaol]

Hi everyone,

I am trying to use EnhancedVolcano in R to plot the results from my rna-seq analysis. However, when I want to use selectLab to label some genes that I want, I only got a weird number labeled on the final plot rather than the gene name...

Here is my code,

raw_data <- read.csv('Directory to raw data .csv file')

mydata <- data.frame(gene = raw_data$gene_name,
             log2_fold_change = raw_data$log2FoldChange,
             adj_p = raw_data$padj) 

EnhancedVolcano(mydata, 
            lab = mydata$gene,
            x = "log2_fold_change",
            y = "adj_p",
            xlim = c(-6,6),
            ylim = c(0,20),
            pointSize = 1,
            selectLab = c('NOX4'))

Here is the what it looks like in the final plot: https://drive.google.com/file/d/17Txw4QEIo1KSgZHJsa-Mn8OLO7ZZ4-Gp/view?usp=sharing

Here is a snippet of the head in my data file: https://drive.google.com/file/d/1yE9ffhmNg98X6wZmOVQZayEVLPKb6i0B/view?usp=sharing.

I tried grepl("NOX4", mydata$gene), and it returned TRUE on the "NOX4" index and FALSE elsewhere.

My code gives me correct results without using the selectLab argument...

Any comments would be helpful.

Thanks in advance

[Mark]

fyi the file you are sharing is locked.

okay got access to the file but it's just a plot... can you post a snippet of your data?

[renhaol]

thanks for pointing that out. It should be fixed!

[Mark]

can you post a snippet of your data?

[renhaol]

Yes, I posted a new screenshot, and added the code to read the file and generate a data frame.

[renhaol]

ok, I added a new screenshot.

[GenoMax]

For future reference when sharing images use: How to add images to a Biostars post

[renhaol]

Thank you, I will follow these in the future.

Answer 1

[renhaol]

ok, problem solved.

For future reference, after I constructed the data frame, the gene name column became "integer" rather than "character" (use typeof() to check. I know, it doesn't look like integers at all...).

I changed my code to the following and it worked as expected.

EnhancedVolcano(mydata, 
        lab = as.character(mydata$gene),
        x = "log2_fold_change",
        y = "adj_p",
        xlim = c(-6,6),
        ylim = c(0,20),
        pointSize = 1,
        selectLab = c('NOX4'))

Helpfully this will help someone else.

[Mark]

thanks for posting the answer! good job on solving it :)

[Kevin Blighe]

Thanks - I developed this package and, indeed, it should be a character vector.

[renhaol]

Thanks! It just did not look obvious to me that my data is not a character vector. I assume it would be some issues from the previous analysis...