Suggestion for circlize package in R for plotting the datasets

Question

[sinha.puja]

Hi All,

I have large data set with methylation values ranging from 0-1. The data contains around 3 million rows of values corresponding to the chromosomal locations. I want to split/sort my data for plot using circos plot. I thought of doing like sorting the value range 0.2-0.4 as hypomethylation category and range 0.7-1.0 as hypermethylation category. However, with this approach the plot again looks cluttered. Another way is that I can split the data into two parts, one part with values lower than median, and another part with values larger than median.

Can anyone suggest me which approach is better or if there are other approaches I can implement in my datasets to make data possible for circos plot using circlize package.

Thanks in advance.

[Kevin Blighe]

Cross-posted on Bioconductor: https://support.bioconductor.org/p/128388/

Answer 1

[Arup Ghosh]

Check this tutorial by the developer of the circlize package. https://jokergoo.github.io/blog/html/segmentation_WGBS_dmr.html

[sinha.puja]

Thank you for the tutorial but still I did not get the appropriate answer.

[Arup Ghosh]

It's hard to provide a better solution without knowing the number(large data is a very ambiguous term) of data sets you are trying accommodate in a plot and how the Circos plot outcome looks with your current configuration.

[sinha.puja]

It's very dense and I don't see the individual dots except two separate bands in the track based on my two range of data sets.

[Kevin Blighe]

sinha.puja, to be fair, if you want help with something like this, then you need to post the exact code that you have been using, and also paste samples of the data that you currently have. You have not even provided a screenshot of what you are seeing on your screen.

[sinha.puja]

The code is below:

circos.initializeWithIdeogram(species = "rn6")
circos.par("track.height"=0.20)
circos.genomicTrackPlotRegion(data = yc,ylim = NULL, numeric.column = 4, 
+                               panel.fun = function(region,value,...) {
+                                 cond <- value[,1] < 0.7
+                        circos.genomicPoints(region[cond,], value[cond,], pch = ".", cex = 0.1,
+                                                      col = "red")
+                      circos.genomicPoints(region[!cond,], value[!cond,], pch = ".", cex = 0.1,
+                                                      col = "blue")
+                               })

The snapshot of image link is below:

Circos Plot

About my data I already described above consisting of 4 columns( Chromosome, start, end and methylation values).

Thanks.