Hi all,
I generated a plot via deeptools2 with the folliwing steps:
BamCoverage to generate a bigwig file from my bam file with default conditions plus --normalizeUsing RPKM
Next, i created a matrix at the regions of interest with computeMatrix and then used plotProfile to visualize this.
Here is my question:
What would be the unit scale on the y-axis?
I assume RPKM? However, I am not sure what underlying computation are done in the steps in between - and I am reluctant in using this. I looked in the literature where they used deeptools2 as well and there is a mix from r/bp, no unit at all, normalized reads, relative fold-enrichment, and seldom RPKM etc etc...
Maybe someone has experience with this? I think people from the deeptools developers are active here as well?
Thank you in advance.
•
written
by
sim.j.baum
The scale would be "RPKM-normalized coverage". There's no further scaling done once you've made your bigWig files.
Thank you. Its a related question, but could probably fill a seperate question: RPKM vs RPGC? Both normalization methods are fine to use in the work stream performed above?
Yes, though I prefer RPGC for everything.
Log in to answer this question.
