Bands with custom colors

[caonetto]

Hi, Thanks for developing the package looks great. I wonder if it’s possible to plot any sort of bands using custom colors? Something like the cytobands but input a bed file with regions and a column with colors associated to the type of region?

Thanks

[bernatgel]

Hi @caonetto

Sure, it's possible to plot any bands as the chromosomes bands. You can do it in two different ways:

• Simulating cytobands: This is not my preferred option and a bit hacky, but you can check https://bernatgel.github.io/karyoploter_tutorial//Tutorial/CustomGenomes/CustomGenomes.html to see how to create a cytobands file and then pass the function a vector of colors. If you want more help on this, ask me.
• Using data.panel: This is my preferred option and I think this is the cleanest and most versatile. You'll need to use data.panel="ideogram" to plot on the ideogram (you can see how to use it at https://bernatgel.github.io/karyoploter_tutorial//Tutorial/DataPanels/DataPanels.html) and combine it with kpPlotRegions. For example, you could plot the chromatin state track seen in the ENCODE example (https://bernatgel.github.io/karyoploter_tutorial//Examples/EncodeEpigenetics/EncodeEpigenetics.html) and plot it on the ideogram adding data.panel="ideogram" to the kpPlotRegions call. If you want to try this option and find problems, ask me here or in biostars and I'll be happy to help.

[caonetto]

Thank you for your response, Is there any option to also plot data directly from lets say a gff file?? How do you upload all the associated data included in a GFF file from the loaded genome and plot densities of different things included in the gff file? Any idea of a tutorial that could be close enough to achieve this?

Thanks you.

[bernatgel]

So far GFF files are not directly supported. You'll need to use another package to convert them into an accepted format.

One possibility is to use rtracklayer::import to import a GFF and then filter the GRanges to obtain the different elements and plot the densities with kpPlotDensity. Something silinar to this (unfinished and untested code):

library(rtracklayer)
gff.data <- import("data.file.gff")
only.exons <- gff.data[gff.data$type=="exon"]
kp <- plotKaryotype()
kpPlotDensity(kp, data=only.exons)

On the other hand, if you have a GFF representing genes, you can call GenomicFeatures::makeTxDbFromGFF and then use the (unsufficiently documented at the moment) functions makeGenesDataFromTxDb and kpPlotGenes (and maybe addGeneNamesand mergeTranscripts) to plot the gene structures.