al2na / methylKit

R package for DNA methylation analysis
https://bioconductor.org/packages/release/bioc/html/methylKit.html
210 stars 96 forks source link

MethylationStats shows one peak #317

Closed gevro closed 7 months ago

gevro commented 7 months ago

Hi, My methylationstats profile doesn't look like what is in the vignette. Is this a computational issue or does this suggest a specific sample issue?

Thanks

Screenshot 2024-03-13 at 8 29 01 PM
alexg9010 commented 7 months ago

Hi @gevro,

if there were no errors when importing your data I assume it is a sample-specific issue. Could you provide further details about the data source and the sample?

You should probably also consult the person who generated the data, to see if this is expected.

Best Alex

gevro commented 7 months ago

It is a bisulfite RRBS WGS sample. Is this simply an indication of low conversion efficiency?

alexg9010 commented 7 months ago

Hi @gevro

The methylation percentage pattern for RRBS is different than the one for WGBS. RRBS targets sites with high CpG density, e.g. CpG islands, which are usually highly methylated. WGBS covers the whole genome including areas with low CpG density, which are usually lowly methylated.

See https://www.illumina.com/science/sequencing-method-explorer/kits-and-arrays/rrbs-seq-scrrbs.html

Best Alex

gevro commented 7 months ago

Sorry, I meant to write it is WGBS. Is this pattern abnormal for WGBS?

alexg9010 commented 7 months ago

Hi @gevro,

This depends on the organism that you are studying. For mammals, the normal pattern would be a bimodal distribution for CpG and very low methylation for non-CpG sites, like in the vignette. On the other hand, for plants, you would expect a different pattern where also non-CpG sites are methylated (see here).

Best Alex

gevro commented 7 months ago

Thanks!