Unite() function in methylkit shows error when destrand = TRUE

[ChandanaRao4]

Hi all,

I am encountering this problem with the unite () function in methylkit when I set the destrand option to TRUE, it gives an error:

Error in unite(methRawList.filt, destrand = TRUE, min.per.group = 2L) :
no base were united. try adjusting 'min.per.group'.

I have checked my files and the num CpGs and that all files have the same naming of Chromosomes and the error still persists.
However, I tried the same with destrand = FALSE and then it works very smoothly.

I am trying to update the pacakage and see if it'll work then.

What could the error be and how do i solve it? Any input would be much appreciated! Thanks in advance :)

[alexg9010]

Hi @ChandanaRao4 ,

This error occurs when merging the methylRawList fails because not enough bases were shared between the samples. I saw from the error message that you have been filtering the methRawList before. Did you check that you retained enough sites?

Could you please provide the code that you used, and if possible, a small sample of your files so we can reproduce the issue? It would also be good if you could tell us a bit more about your setup, like how many samples and how many samples per group you are working with.

Best,
Alex

[ChandanaRao4]

Hi Alex, Thanks so much for your response! I filtered out the high coverage sites that can be problematic for the downstream analysis: 1. So I started by reading all my coverage files into a methylraw list with the following function methRawList <- methRead(location = files.all, sample.id = sample.id.list, assembly = "hg38", treatment = ordered.metadata$diagnosis_group, pipeline = "bismarkCoverage", context = "CpG", mincov = 10) Then I checked the coverage statistcs and methylation perecentage and some other basic QC. 1. Then I filtered our sites that have very high coverage (poetnatil PCR duplicates) methRawList.filt <- filterByCoverage(methRawList, lo.count=NULL, lo.perc=NULL, hi.count=NULL, hi.perc=99.9) 1. Then I used the unite function to merge the data. And it only works when destrand is set to FALSE meth.all.filt <- unite(methRawList.filt, destrand=FALSE, min.per.group = 1L) No matter which number I set the min.per.group, it slways says cannot unite bases * I have verified the structure of my methylRawList object, I have checked the num CpGs per sample. There was only one sample with low num CpGs but the rest of them look fine. The library prep was done with EMseq and not the traditional bisulfite conversion. * I have 28 samples in total, and they were run on two lanes of NovseqX plus and the facility performed the data preprocessing for us. We asked them the merge the data form the two lanes and also to do a CpG merge to increase the coverage. I am afraid I cannot share the data at this point. I will need to check before doing so. Appreciate your help and input! Thanks so much in advance 😊 Best, Chandana Rao |PhD Student- Jagodic group Department of Clinical Neuroscience, Center for Molecular Medicine (CMM), Visiongatan 18, L8:04, 17164, Karolinska Institutet, Stockholm, Sweden. From: Alexander Blume ***@***.***> Date: Friday, 1 August 2025 at 10:23 To: al2na/methylKit ***@***.***> Cc: Chandana Rao Prakash ***@***.***>, Mention ***@***.***> Subject: Re: [al2na/methylKit] Unite() function in methylkit shows error when destrand = TRUE (Issue #362) You don't often get email from ***@***.*** Learn why this is important<https://aka.ms/LearnAboutSenderIdentification> [Image removed by sender.]alexg9010 left a comment (al2na/methylKit#362)<#362 (comment)> Hi @ChandanaRao4<https://github.com/ChandanaRao4> , This error occurs when merging the methylRawList fails because not enough bases were shared between the samples. I saw from the error message that you have been filtering the methRawList before. Did you check that you retained enough sites? Could you please provide the code that you used, and if possible, a small sample of your files so we can reproduce the issue? It would also be good if you could tell us a bit more about your setup, like how many samples and how many samples per group you are working with. Best, Alex — Reply to this email directly, view it on GitHub<#362 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/BPBPH7FE4JPSBIPWQOSIHWD3LMPYLAVCNFSM6AAAAACCZMQETSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCNBTGY4TOMJSGU>. You are receiving this because you were mentioned.Message ID: ***@***.***> När du skickar e-post till Karolinska Institutet (KI) innebär detta att KI kommer att behandla dina personuppgifter. Här finns information om hur KI behandlar personuppgifter<https://ki.se/om-ki/integritetsskyddspolicy>. Sending email to Karolinska Institutet (KI) will result in KI processing your personal data. You can read more about KI’s processing of personal data here<https://staff.ki.se/data-protection-policy>.

[alexg9010]

Hi @ChandanaRao4,

You mentioned that your facility already

[..] performed a CpG merge to increase the coverage. [..]

I assume that this was done using the --merge_CpG flag and likely caused the error during destranded uniting. My suggestion is to skip the destranding since you already merged the signal accross those dinucleotides.

Unfortunately the destranding did not cause an error, thus the merging silently failed, so we need to dig a bit deeper here.
Please follow up with your stakholders, but it would be great if you could share 100-200 lines of the bismarkCoverage files that you used, such that we can generate a tiny test set that reproduces the issue.

Best,
Alex