I have enzymatic methyl-seq (EM-seq) data, is it possible to use it to look for variants/mutations since its whole genome sequencing? The EM-seq data is composed of two groups, exposed and control.

yup, I think it should be mentioned in the docs, no?

Oh yes, but not for methylation calling, but rather nucleotide mutations

ah right, sorry - read too quickly and missed that 😅

ah, will do!

Say if I made that up and you don't find anything 😅

still diggin’ hehe

hah, 6 year old GitHub issues, my favourite.. 😆

Hi Andrey, thanks for that question. In short, I don't think the methylseq nf.core pipeline is currently set up to do variant calling. The truth is that variant calling from bisulfite and EM-seq data is hard, and the quality will be a lot worse than with WGS data, for all I know. There are tools out there, e.g. "biscuit" or "gemBS", which can do it, but expect a relatively high error rate. For TAPS and 5Base data, we are about to release rastair v2 which does variant calling with something like 99.8% accuracy for 20x depth data. It would be nice to include that into the methylseq pipeline, but it will not work for EM-seq I'm afraid

Thanks for the feedback guys!

One option is to mask converted bases before variant calling. Here is the poster by NEB describing the details: https://www.neb.com/en/-/media/posters/document/germline-variant-calling-with-em-s[…]ef3f64acc88d94&sc_lang=en&hash=7243893FF857BCA472B950BC3F904038

Maybe also worth mentioning, there are a number of tools out there attempting to perform variant calling in BS-seq data leveraging the influence of the opposing strand on C>T SNPs, such as: Bis-SNP, MethylExtract, BS-SNPer or CGmapTools. Reading one ore more of their publications should give you an answer (and they should all work with EM-seq data)