Unable to alter margins in base R boxplot

[kubu4]

Firstly, here's the R Project I'm working with:

https://github.com/epigeneticstoocean/2018_L18-adult-methylation/tree/main/code/ballgown_diff-isoform-expression

If you download that entire directory, you'll be able to reproduce everything. NOTE: You'll need to install ballgown from BioConductor (good luck!)...

So, if I generate a box plot using the following code, here's what it looks like:

# Pull all transcript expression values stored in FPKM measurement from ballgown object
fpkm <- texpr(bg, meas = "FPKM")

# Log transform data and add 1 to all values to artificually prevent log2(0).
fpkm <-log2(fpkm+1)

# Set plot colors for F/M
myColors <- ifelse(phenotype_table$Sex=="F", rgb(0.8,0.1,0.3,0.6),
                   ifelse(phenotype_table$Sex=="M", rgb(0.1,0.1,0.7,0.5),
                          "grey90")
)

# Generate box plot.
boxplot(fpkm, col = myColors, las=2, ylab='log2(FPKM+1)')

---

Now, if I want to edit the margins so that I can see the full text on the x-axis, I add this line and here's how the plot looks now:

# Set box plot margins (extreme values to show no effect on boxplot)
par(mar=c(100,100,100,100))
# Generate box plot.
boxplot(fpkm, col = myColors, las=2, ylab='log2(FPKM+1)')

---

So, anyone have any ideas on how to resolve this?

The alternative (and, arguably better approach), is to use ggplot2, but I'm not sure where to even start. Some of my feeble attempts have led to nothing but errors (Error: datamust be a data frame, or other object coercible byfortify(), not a numeric vector.) or (if I convert the fpkm object to a data frame) lack of understanding on how to use the aes() aspect of ggplot().

Would appreciate any/all assistance!

[kubu4]

And, here's a link to a tab-delimited version of the ballgown object I am working with:

https://github.com/epigeneticstoocean/2018_L18-adult-methylation/blob/main/code/ballgown_diff-isoform-expression/analyses/fpkm_all_samples.tab

[sr320]

Done in tidyverse

https://github.com/epigeneticstoocean/2018_L18-adult-methylation/blob/main/code/05-ballgown-boxplots.md

[kubu4]

Just adding written description for posterity.

Didn't resolve the actual base R boxplot margin issue.

Instead, went with ggplot(). This also required using pivot_longer() to rotate the data frame for R to be happy with it.