error in rainfallPlot "cannot mix zero-length and non-zero-length coordinates"

[bkinnersley]

Hello,

Thank you for a very useful package! I am interested in running the rainfallPlot function to detect potential regions of kataegis. This runs fine for most tumours I have tried it on, e.g. running example command:

library(maftools)

input_maf = 'sample_id.GRCh38.PASS.MAF.gz'

input = read.maf(maf = input_maf)

rainfallPlot(maf = input, detectChangePoints = TRUE, ref.build = "hg38", savePlot = TRUE)

However for a very small number I get the following error message:

Kataegis detected at:
Chromosome Start_Position End_Position nMuts Avg_intermutation_dist Size Tumor_Sample_Barcode C>T T>C

1. 1 630110 634494 8 626.2857 4384 sample_id 3 5

Error in arrows(x0 = maf.cpt$Start_Position_updated, y0 = 0, x1 = maf.cpt$Start_Position_updated, :
cannot mix zero-length and non-zero-length coordinates
Calls: rainfallPlot -> arrows
Execution halted

The errors always seem to occur if a potential region of kataegis is detected right at the start of chromosome 1 (with Start_Position and End_Position very similar and often the same as those in the example above)

Thanks very much!

[ShixiangWang]

Are you using the latest version of maftools from GitHub? Could you also post the data for debugging?

[bkinnersley]

Hello,

Thanks for the reply! Attached is my sessionInfo() - I used the latest version of maftools (v2.5.10)

It is not possible to upload a whole MAF but here are the mutations in one MAF that seem to have caused the error:

Thanks very much

[ShixiangWang]

If you could provide some mutation records to reproduce the error, you should also attach a file so we can use it for import.

[bkinnersley]

Hello,

Unfortunately I cannot upload the whole MAF, but have created an initial subset that reproduces the issue, hope this helps

example.zip

Thanks very much

[ShixiangWang]

@bkinnersley I have submitted a PR #629 to fix this issue. You can install the latest version from GitHub after @PoisonAlien accept the PR.

[PoisonAlien]

Hi @bkinnersley
Thanks for opening the issue. I have pushed a fix using PR #629
Please try it and feel free to reopen the issue if necessary.

[bkinnersley]

Thanks very much that fixed it