Arranging Freq and Count Plot by same sample

[profsprofs1991]

Hi,

I have been exploring the package and so far it has been very useful (specially since it is my first time working with mutational signatures). I generated the freq/Count of a number of selected COSMIC signatures in my cohort and displayed in the plot below.

I noticed the msigPlot function arranges each sample by descending count, but when the frequency is plotted I think it end up being arranged by a different factor. Is there a way to ensure both plots are arranged in the same fashion (or is it even relevant?).

I also wanted to ask if you have any documentation on how to compare specific signatures between two cohorts of patients.

Thanks a lot in advance,

Mateo Mejia

[dami82]

Hi Mateo,

Thanks for using mutSignatures. Yes, that may be very relevant. For example, we built a similar visualization in Sci Rep. 2020; 10: 18217. Fig5. You raised a valid point, mSigPlot() cannot generate two matched plots at the moment. However, you can build them using ggplot2. The trick is to coerce both signature exposure objects (absolute counts; relative/freq) to data.frame, melt them, and then re-factor the 'Sample' columns in both columns so that their levels match. I attached a vignette that includes an example and shows how to do just that: mutSignature_test_v230316.pdf

About your second question. Signatures (but NOT signature exposures) can be compared using the matchSignatures() function. More details are available here:

• https://doi.org/10.1038/s41598-020-75062-0
• ?matchSignatures()

I hope you find this useful.