local-methylation-profile

name	local-methylation-profile
description	This skill analyzes the local DNA methylation profiles around target genomic regions provide by user. Use this skill when you want to vasulize the average methylation profile around target regions (e.g. TSS, CTCF peak or other target regions).

Local Methylation Profile Analysis

Overview

• Always prompt user for which columns in the BED files are methylation fraction/percent. Never decide by yourself.
• Generat profile: Bin methylation around regions (±flank, fixed bin size), aggregate mean±SE.
• Visualize: Plot mean profile with ribbon and center line.

---

Inputs & Outputs

Inputs

methylation.bed
target_regions.bed

Outputs

local_methyl_profile/
  stats/
    CpG_around_target.tsv
  plots/
    CpG_around_target.pdf
  temp/
    ... # other temp file generated

---

Decision Tree

Step 1: Preprocess input → 5-column BED (for methylKit), and 3-column BED (for target regions)

awk -F'\t' 'BEGIN {OFS="\t"} {print $1, $2, $3, $<i_methylation>}, $<i_coverage>}' methylation.bed # n is provide by user, *100 if is fraction 
awk -F'\t' 'BEGIN {OFS="\t"} {print $1, $2, $3}' target_regions.bed

---

Step 2: Build methylation profiles around regions

Call:

• mcp__methyl-tools__build_local_methylation_profile

with:

• methyl_bed_path: 5-column BED-like file from preprocess_methylation.
• regions_bed_path: 3-column BED-like file from preprocess_regions.
• output_profile_tsv_path: path for aggregated profile table (TSV).
• flank_size: flank size in bp around region center (default 2000).
• bin_size: bin size in bp (default 50).
• min_coverage: minimum coverage threshold for CpGs (default 10).

---

Step 3: Visualization

Call:

• mcp__methyl-tools__plot_profile

with:

• profile_tsv_path: TSV from build_methylation_profile.
• output_plot_path: output figure path (PNG/PDF; format inferred from extension).
• title: plot title (optional).

---

Parameter Guidelines

Context	Flank	Bin	Min cov
TF peaks	±2 kb	50bp	10x
Promoters	±1 kb	50bp	10x
Enhancers	±5 kb	100bp	5x
Motifs	±0.5kb	10–20	10x

Notes

• Snippets are usage hints and must be adapted to your paths and column indices.