How to recreate this dot plot?

Question

[Ron]

Hi all,

Does somebody know which package can be used for plots like this?

The dot plots illustrates the number of detected fusion transcripts per megabase per sample normalized by the sequencing coverage. Tumor types were sorted according to the fraction of samples with fusions.

I am trying to recreate this plot from the gene fusions paper. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4468049/

The data is present here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4468049/bin/NIHMS632238-supplement-2.xlsx (Table S1) and using only columns "Cancer, FileName,FusionPair"

I have tried a couple of options from ggplot package but these are different plots. http://www.sthda.com/english/wiki/ggplot2-dot-plot-quick-start-guide-r-software-and-data-visualization

Thanks

Ron

[Ram]

This post needs a better title and more relevant tags. At first sight, this post can likely be dismissed as a pure R question but you're looking to recreate a certain type of plot in R depicting gene fusion events across TCGA cancer datasets. Please add such information in there to make your post more relevant to bioinformatics.

[russhh]

It is possible to do them in ggplot2; I'll try and find the code. You setup your plot to use the disease-factor as x and number of mutations as y; I think you need an ordering variable nested within each disease-value (from -0.5 to 0.5) that dictates the x-offset for each point.

[Ram]

Or a facet, maybe? (Unless I'm missing something)

[russhh]

facets would generate several graphs, perfectly fine for exploration. Here they want a single graph.

[Ram]

It still produces one graphical entity. By several graphs, do you mean repeat the axes several times? That can be controlled.

[russhh]

Could you post some example data, please?

[Ron]

I posted the link to the data and the paper.

[zx8754]

Related post: How to generate a stripplot like this in Python or R?

Answer 1

[zx8754]

Here is the start:

library(ggplot2)

#dummy data
mydat <- mtcars[ , c("cyl", "mpg")]
mydat$Cancer <- paste0("cancer_", mydat$cyl)

#Order by cancer and mpg (number of fusions per megabase)
mydat <- mydat[ order(mydat$Cancer, mydat$mpg), ]
mydat$x <- seq(nrow(mydat))

# make pretty x labels one per cancer
mydat$xlabel <- c(1, diff(as.numeric(as.factor(mydat$Cancer))))
mydat$xlabel <- ifelse(mydat$xlabel == 1, mydat$Cancer, NA)

# plot
ggplot(mydat, aes(x, mpg, col = Cancer)) +
  geom_point() +
  scale_x_continuous(breaks = which(!is.na(mydat$xlabel)), 
                     labels = mydat$xlabel[ !is.na(mydat$xlabel)])

[cpad0112]

I think this would answer this post : C: How to generate a stripplot like this in Python or R?