MethScope-MRMP

In MethScope: Ultra-Fast Analysis of Sparse DNA Methylome via Recurrent Pattern Encoding

options(device = "png")
knitr::opts_chunk$set(
  fig.ext = "png",
  collapse = TRUE,
  comment = "#>"
)

Generate MRMP reference

We provided three reference MRMPs definitions, and if you want to create your own MRMPs definitions, please follow the following tutorial.

1. Download and install YAME (https://zhou-lab.github.io/YAME/)
2. Process you methylation reference file into .cg file
3. run the following yame rowop code to binarize your input file (default threshold is at 0.5, use -b to adjust threshold if needed)

```{bash, eval=FALSE} yame rowop example.cg -o binstring > example_binstring

4. Generate the reference definition (default min CG per MRMP is 50, adjust the threshold if needed with min_CG)

```r
GenerateReference(example_binstring)

1. Pack the generated pattern.txt into .cm file which can be directly used as reference_pattern in our GenerateInput function.

```{bash, eval=FALSE} yame pack -f s patterns.txt patterns.cm

## Generate MRMPs reference from cell type pseudobulks

It might be useful to generate the MRMPs reference from cell type pseudobulks if the cell type label is available. 

1. Obtain a txt file that contains sample names belong to oen cell type that you want to merge together from the cell type label.
2. Generate the pseudobulk profile using yame subset and rowop

```{bash, eval=FALSE}
yame subset -l samples.txt example.cg | yame rowop -o musum - example_pseudobulk.cg

Please check our full code GenerateReference.sh in our analysis folder (https://github.com/zhou-lab/MethScope) as a reference for in parallel process multiple cell types. After obtaining the pseudobulk profile, it will be the same steps as above to obtain the MRMPs definition file.

MethScope documentation built on Feb. 27, 2026, 1:08 a.m.