nloyfer / wgbs_tools

tools for working with Bisulfite Sequencing data while preserving reads intrinsic dependencies
Other
134 stars 37 forks source link

define the cell type-specific markers, as defined in the Nature paper #83

Open gyuhee0928 opened 3 months ago

gyuhee0928 commented 3 months ago

I encountered an issue while trying to define the cell type-specific markers, as defined in the Nature paper (Loyfer, N., Magenheim, J., Peretz, A. et al. A DNA methylation atlas of normal human cell types. Nature 613, 355–364 (2023).), using find markers.

I'm attempting to replicate the methods form the Nature paper. To do this, I downloaded all the beta and pat files from GSE186458. I tried to define the top 1000 cell type-specific markers to replicate markers listed in supplementary table 4c.

Herer are the commands I used:

wgbstools segment --beta_file total_beta_file --min_cpg 5 --max_bp 5000 -@ 85 -o NMA.blocks.bed

wgbstools find_markers --blocks_path NMA.blocks.bed --groups_file total_beta_file.group.csv --beta_list_file total_beta_file --min_cpg 5 --min_bp 10 --max_bp 1500 -@ 85 --only_hypo -c 25 --top 1000 -o NMA.dmr.U.top1000

However, I noticed there was very little overlap between the marker I defined and those in supple table 4c. So, I checked all the dataset i used. but there was nothing wrong

Could you please help me identify what might be causing this discrepancy?