Hi all,
In our lab we do a lot of RRBS sequencing...
# methylseqe
Edoardo Giuili
12/05/2025, 1:30 PMHi all,
In our lab we do a lot of RRBS sequencing (on both tumor tissues and liquid biopsy samples). As you might know, computing sequencing saturation on RRBS samples is not possible with tools like Preseq since most of the reads are labeled as PCR duplicates. That's why we developed a tool, methurator (https://github.com/VIBTOBIlab/methurator), to estimate sequencing saturation on RRBS samples. It's already published on PyPi and available via Bioconda and Biocontainers. Would you guys be interested in having it in the methylseq pipeline when the --rrbs flag is activated? If yes I can make a PR for future releases? Let me know what you think about it, any feedbacks are more than welcome :)
🎉 2
p
Phil Ewels
12/05/2025, 7:28 PMHi @Edoardo Giuili! The main output is this graph, right? (Thanks for adding it to the readme 🙏🏻 )
Phil Ewels
12/05/2025, 7:30 PMWhat do you think about adding support to MultiQC for the outputs? Then the plots would show up in the already-existing report at the end of the pipeline, rather than requiring the end user to open separate files.
Phil Ewels
12/05/2025, 7:31 PMIt's awkward to have the line change unfortunately, that's the only tricky bit. But I don't think that it loses all that much without it? And could potentially have it as a tabbed double-plot: predicted and observed.
Phil Ewels
12/05/2025, 7:33 PMCould also have a single figure like "% asymptote" or "% theoretical maximum" which could go into the General Stats table at the top 🤔
e
Edoardo Giuili
12/11/2025, 10:23 PMHi @Phil Ewels, thanks! And yes, sounds good to me to add support to MultiQC.
I've incorporated the changes you suggested for the plot. What do you think? In the html file you can also visualize stats for each point (observed or predicted).
Edoardo Giuili
12/11/2025, 10:23 PMScreenshot 2025-12-11 at 23.20.23.png
p
Phil Ewels
12/12/2025, 6:55 AMAh sorry, for the line change I meant specifically if plotting within MultiQC
Phil Ewels
12/12/2025, 6:56 AMI think you can have both - no reason not to also have the
methurator plot👍 1
Phil Ewels
12/12/2025, 8:28 AMSorry, I started poking around the code a little, thinking about making a MultiQC module and ended up making a bunch of PRs 🙄 I can't help myself 😆
Phil Ewels
12/12/2025, 8:28 AMFeel free to ignore / change as you wish
Phil Ewels
12/12/2025, 9:15 AMok I went totally overboard 😂 Sorry, I was having fun / procrastinating.
🎉 1
e
Edoardo Giuili
03/31/2026, 5:22 PMSo, methurator is now capable of working with even larger samples (such as WGBS or EMseq), making it a useful addition to methylseq for assessing CpG-level sequencing saturation. We’ve submitted it to JOSS, but you can find a LinkedIn post for now here to get a better idea of how it works, and here for the documentation.
I’d really appreciate your feedback—let me know what you think. If there’s interest in integrating it into methylseq, I can create a new nf-core module and open a PR.
@Phil Ewels: the latest version (2.1.1) isn’t yet supported by MultiQC, so I’ll wait for that PR to be approved before opening one for methylseq.