Closed kaanokay closed 1 year ago
Than @kaanokay for the issue, modbams and bedmethyls don't currently work with NanoMethViz, which assume per-read data with coordinates and modification probabilities. This package currently outsources DMR detection to packages like edgeR, bsseq and DSS.
It's very useful to have the sample data, I think it can be converted into the format used in bsseq and DSS. Both those packages use bsseq objects, I'm not sure if they can import directly from bedmethyl, but it should be feasible and useful.
Hi @Shians,
Thank you so much for your reply. Its so helpful, I'll try analyze my data using bsseq.
Best.
Hi @Shians again,
I was wondering what input should be DMR analysis. It should be mod column/mod colum + canon column or; mod colum/coverage column for methylation calculation? I'm not sure about it. Do you have thoughts on it?
Thanks!
I think that will depend on what DMR analysis tool you are using. For example bsseq wants the following for their objects:
Hi,
That's helpful. Thank you!
Bw.
Dear @Shians,
I'd like to thank you for developing this tool.
I have couple of questions about how to process Remora output for DMR analysis.
I have two groups: one of them is mutant, another one is control group. I used modbam2bed tool to get bedmethyl file from Remora bam files. I attached them in this post.
Could you help me how to find DMR regions between mutant and control groups using bed files through your R package? As I understand, your package supports Megalodon and Nanopolish but does not support Remora output. Btw, those groups are Mus musculus samples. I don't know what kind of input structure that I can produce as input for your package.
Best.
control.txt mutant.txt