dpryan79/MethylDackel

Chromosomes effected by methylation

emiliomastriani opened this issue · 4 comments

Dear all,
I have troubles to understand same data from the results. I used methyldackel as methyl extract task inside the nf-core/methylseq/bwa-meth pipeline, and it worked perfectly. I have the following open questions:

  1. In the methyldackel file, there is a txt report containing a table wherever indicated indicated strand , read , position, nmethylated , nunmethylated, as reported below. For n methylated i saw (for example) 33 for position 1096 , it means 33 methylated sites identified if i'm right. However for the position 1096 i don't know referring to which chromosome. Please, can you tell me how to connect the methylated region to its chromosome?

Strand Read Position nMethylated nUnmethylated
OT 1 1 742 8
OT 2 1 1234 7
OT 1 2 1299 22
OT 2 2 1195 19
OT 1 3 1270 28
OT 2 3 1168 9
OT 1 4 1096 33
OT 2 4 1102 23

  1. At the end of its execution, methyldackel produced 2 svf pictures reporting the Position along mapped read/CpG Methylation % for both strand OT/OB respectively, as reported at https://github.com/dpryan79/MethylDackel#Methylation bias plotting and correction. Also in this case, I don't understand to which chromosome every position refers.

Thank you for you support.
Dr. Mastriani Emilio

  1. That appears to be the methylation bias output, which is the text version of plots like https://camo.githubusercontent.com/95b03c9e48fbc805138562571b7a326c5636ddce29b2cc62ee57e4b366040982/68747470733a2f2f7261776769742e636f6d2f64707279616e37392f4d657468796c4461636b656c2f6d61737465722f6578616d706c655f4f542e737667 I assume nf-core put the actual methylation metrics elsewhere.
  2. That is all of the chromosomes, which you need to look at to know if there are portions of your reads that should be ignored.

Ask the nf-core folks where to find the actual per-base methylation metrics. They're normally named something_CpG.bedGraph, but they may have changed that.

Well. I know the file you refer. In my test it's named 5_BKDL220006759-1a.sorted.markDups_CpG.bedGraph and looks like the following:

track type="bedGraph" description="5_BKDL220006759-1a.sorted.markDups CpG methylation levels"
chr1 16555 16556 100 1 0
chr1 16570 16571 100 1 0
chr1 16617 16618 100 1 0
chr1 16619 16620 100 1 0
chr1 16634 16635 100 1 0
chr1 16646 16647 100 1 0
chr1 16665 16666 100 1 0
chr1 16667 16668 100 1 0
chr1 17378 17379 100 1 0

So, if I well understand, this is the file containing the methylation metrics needed for further analysis.
Thank you

Correct, glad things got cleared up!