Hello everyone, I'm currently running a MDS / PCA analysis in the context of a RNASeq experiment as suggested in this RNASeq edge quasi-likelihood pipeline and particularly through the R function plotMDS.

I'm working on a particularly complex dataset, with almost one hundred samples, and the samples appear to cluster in 4 different groups, quite distant on the plot. As a result, while inserted in a document (let's say, a Word page, but possibly also a scientific paper), the symbols appear very small.

Of course, I could provide a 90° rotated plot, but I was wondering if there is a way to make symbols bigger, so the resulting plot can be read better?

Thank you in advance for your help.

The symbol size is controlled the same way for plotMDS() as for any plot function in the stats package of R. The symbol size isn't determined by the function itself but rather by the graphic functions of R.

How exactly did you save the plot? Did you write is as a pdf, or as a Windows meta file, or as something else? What R function did you use to save the plot?

plotMDS gives you complete control over symbol size and other plotting parameters, but there is actually no need to manipulate any of those parameters. The reason the symbols are so small in the plot is that you created the plot with a very large width and height. If you simply make the virtual width and height of the plot window smaller, then the symbols and x- and y- labels etc will all become automatically relatively larger. If you are saving the plot from an interactive window, just make the window smaller.