Guidance needed for interpreting ONT methylation DMR results and methylartist inputs

[Jyoti-Mridha666]

Hi Team,

I hope you are doing well. I am new to methylation analysis and currently learning to work with ONT Epi2Me Human Variation and methylartist/modkit tools. I would sincerely appreciate your guidance in interpreting my methylation results and understanding the appropriate next steps.

Background and workflow

I ran the ONT Epi2Me Human Variation workflow using the following command:

nextflow run /mnt/c/Users/shibi/epi2melabs/workflows/epi2me-labs/wf-human-variation
-profile standard
-c /mnt/c/Users/shibi/epi2melabs/workflows/epi2me-labs/wf-human-variation/nextflow.config
--bam merged.rebuilt.sorted.bam
--ref GCA_000001405.15_GRCh38_no_alt_analysis_set.fa
--sample_name 20251015SEQ_human_variation
--override_basecaller_cfg dna_r10.4.1_e8.2_400bps_sup@v3.5.1
--threads 14
--bam_min_coverage 5
--phased
--snp
--GVCF
--sv
--cnv
--mod
--str
--out_dir with_methylation_basecalling/

From this run, I successfully obtained results related to SNPs, SVs, STRs, CNVs, and methylation (--mod).

Methylation-related output files

The methylation-specific files generated by the Human Variation workflow are:

(1) 20251015SEQ_human_variation.wf_mods.1.bedmethyl
(2) 20251015SEQ_human_variation.wf_mods.2.bedmethyl
(3) 20251015SEQ_human_variation.wf_mods.ungrouped.bedmethyl

=> Since my primary interest is in methylation patterns, I proceeded with downstream analysis using modkit.

Differential methylation analysis using modkit

I performed haplotype-based DMR analysis using the following command:

modkit dmr pair
-a 20251015SEQ_human_variation.wf_mods.1.bedmethyl.gz
-b 20251015SEQ_human_variation.wf_mods.2.bedmethyl.gz
--ref GCA_000001405.15_GRCh38_no_alt_analysis_set.fa
--base C
--threads 22
-o haplotype_DMRs.bed

=> This is the result for my differentially methylated regions:

base) ubuntu@EpistemeMK1B:/mnt/c/Users/shibi/OneDrive/Desktop/SEQUENCING_IDs/1_Blood_Sample_Result_Methylation_Human_Variation/1_Human_Variantion_Analysis_result_files/modkit$ head haplotype_DMRs.bed
chr1 12304 12305 . -0.4795730802618845 + h:0,m:1 1 h:0,m:2 2 h:0.00,m:100.00 h:0.00,m:100.00 1 1 1 0 0 -2.4004558381776544 2.4004558381776544
chr1 12330 12331 . -0.4795730802618845 + h:0,m:1 1 h:0,m:2 2 h:0.00,m:100.00 h:0.00,m:100.00 1 1 1 0 0 -2.4004558381776544 2.4004558381776544
chr1 12354 12355 . -0.4338645826298626 + h:0,m:1 1 h:0,m:1 1 h:0.00,m:100.00 h:0.00,m:100.00 1 1 1 0 0 -2.7718076486993555 2.7718076486993555
chr1 12425 12426 . -0.4795730802618845 + h:0,m:1 1 h:0,m:2 2 h:0.00,m:100.00 h:0.00,m:100.00 1 1 1 0 0 -2.4004558381776544 2.400455838177654

=> This produced the haplotype_DMRs.bed file. I have reviewed the output, but as I am completely new to methylation studies, I am finding it difficult to clearly understand:

• How to correctly interpret these DMR results
• Which columns/parameters are most important
• What biological conclusions or end results I should reasonably expect
• How these results can be used meaningfully in downstream analysis

Although I have referred to several papers and documentation, I still have several conceptual questions, and expert guidance would greatly help me gain clarity and confidence.

=> Files currently available
From my analysis so far, I have the following files:

=> i have this results along with me based on my previous run

• After methylation mode basecalling at high accuracy mode (This i used for human variation workflow):
  (1) merged.rebuilt.sorted.bam

• Directly after methylation mode --mod (Epi2me Human variation pipeline):
  (1) 20251015SEQ_human_variation.wf_mods.1.bedmethyl
  (2) 20251015SEQ_human_variation.wf_mods.2.bedmethyl
  (3) 20251015SEQ_human_variation.wf_mods.ungrouped.bedmethyl
  (4) 20251015SEQ_human_variation.haplotagged.cram
  (5) 20251015SEQ_human_variation.haplotagged.cram.crai
  (6) 20251015SEQ_human_variation.mosdepth.global.dist
  (7) 20251015SEQ_human_variation.mosdepth.summary
  (8) 20251015SEQ_human_variation.wf_mods.ungrouped.bedmethyl

• modkit result file:
  (1) haplotype_DMRs.bed

###################

Methylartist visualization

I am also interested in visualizing methylation patterns using methylartist and have been following the documentation from:
https://github.com/adamewing/methylartist

However, I am currently stuck at the very first step and would appreciate clarification regarding the required input format. For example, in the following command:

methylartist db-nanopolish -m MCF7_ATCC_REP1.nanopolish.tsv.gz -d MCF7_ATCC.nanopolish.db

I am unclear about:

• What exact input file is expected
• Whether my existing .bedmethyl or BAM/CRAM files can be used directly
• If additional processing or conversion is required for methylartist
• Request for guidance

I would be very grateful if you could kindly advise on:

• How to interpret DMR results and key parameters to focus on
• Recommended next steps after obtaining haplotype-based DMRs
• Best practices for methylation visualization and biological interpretation
• How to properly prepare inputs for methylartist in my case

Any suggestions, explanations, or pointers to relevant resources would be extremely helpful and appreciated.

Thank you very much for your time and support.

Kind regards,
Jyoti

[ArtRand]

Hello @Jyoti-Mridha666,

Thanks for your question, getting the data prepared and organized is certainly a good first step.

It's hard for me to give you specific advice without knowing anything about your biological question. One process you may try is using --segment in your DMR pair command docs here. You could then look for regions that are identified as different and see if they intersect/overlap with some regions you're interested in studying. As for visualization, I recommend converting the .bedmethyl into a bigWig file docs here then viewing on a browser - again looking at regions you're studying.

I can't advise you on how to operate methylartist. It looks like you've opened an issue on that project so hopefully they can help.