Graphics issue in plotFusion and plotFusionReads about chimeraviz

[stianlagstad]

The first issue is that there's not enough rom on the screen to actually show you all the nucleotides. You can limit the amount of nucleotides with the nucleotideAmount parameter - that'll let you experiment a bit and find the limit. Try setting it to 10 to start, and also set showAllNucleotides to FALSE.

plotFusionReads(fusion = fusion, showAllNucleotides = FALSE, nucleotideAmount = 10)

---

I reproduced the second issue this morning on a mac at work. For some reason it draws the arc outside of the actual plot. But it seems to be a mac-specific problem. I'm not able to reproduce the error with my home laptop running Ubuntu. This code:

library(chimeraviz)
if(!exists("defuse833ke"))
  defuse833ke <- system.file(
    "extdata",
    "defuse_833ke_results.filtered.tsv",
    package = "chimeraviz")

# Then load the fusion events
fusions <- importDefuse(defuse833ke, "hg19", 1)
fusion <- getFusionById(fusions, 5267)
if(!exists("edbSqliteFile"))
  edbSqliteFile <- system.file(
    "extdata",
    "Homo_sapiens.GRCh37.74.sqlite",
    package="chimeraviz")
# Then load it
edb <- ensembldb::EnsDb(edbSqliteFile)

if(!exists("fusion5267and11759reads"))
  fusion5267and11759reads <- system.file(
    "extdata",
    "fusion5267and11759reads.bam",
    package = "chimeraviz")

if(!exists("bamfile5267")) {
  bamfile5267 <- system.file(
    "extdata",
    "5267readsAligned.bam",
    package="chimeraviz")
}

plotFusion(
  fusion = fusion,
  bamfile = bamfile5267,
  edb = edb,
  nonUCSC = TRUE,
  reduceTranscripts = TRUE)

Produces this plot for me:

Edit: Back on the mac, that same code produces this plot:

I'll look into this as soon as I can!

[stianlagstad]

Just to show a sign of life: So far I'm clueless. In the plotFusion function I use Gviz::coords() to fetch the coordinates of where the exons are plotted in the picture. I use these to decide where to draw the red arc. For some reason the y coordinates those code lines returns on a mac are off. I don't know why yet. I don't really think it's a Gviz issue.

Relevant chimeraviz code from plotFusion.R:

    # Get coordinates for the exons in the plot
    exonCoordinates1 <- Gviz::coords(res1$GeneRegionTrack)
    exonCoordinates2 <- Gviz::coords(res2$GeneRegionTrack)

[stianlagstad]

After getting some help on the bioc-devel list I got a bit further. The arc turns up where it's supposed to when I use another graphics device. Could you try that as well? I.e. try:

# Load data and example fusion event
defuse833ke <- system.file(
  "extdata",
  "defuse_833ke_results.filtered.tsv",
  package="chimeraviz")
fusions <- importDefuse(defuse833ke, "hg19", 1)
fusion <- getFusionById(fusions, 5267)
# Load edb
edbSqliteFile <- system.file(
  "extdata",
  "Homo_sapiens.GRCh37.74.sqlite",
  package="chimeraviz")
edb <- ensembldb::EnsDb(edbSqliteFile)
# bamfile with reads in the regions of this fusion event
bamfile5267 <- system.file(
  "extdata",
  "fusion5267and11759reads.bam",
  package="chimeraviz")
# Temporary file to store the plot
pngFilename <- "testplot.png"
# Open device
png(pngFilename, width = 1000, height = 750)
# Plot!
plotFusion(
  fusion = fusion,
  bamfile = bamfile5267,
  edb = edb,
  nonUCSC = TRUE)
# Close device
dev.off()

and then see what testplot.png looks like?

[komalsrathi]

Yes this works for me -- thanks! I am testing the entire code now and will update you on it.

Thanks a lot!

[stianlagstad]

Great! I'll look more into the default graphics driver on mac and try to figure out the issue. But I'll close this :)