Skip to contents

Parse a Megalodon per-read methylation file into a tidy per-site data frame

Source: R/parse_megalodon.R
dot-parseMegalodon.Rd

Reads a Megalodon per-read modification output file, aggregates per-read calls to per-site beta values and coverage, and returns a tidy data frame compatible with the commaData constructor. This is an internal function called when caller = "megalodon".

Usage

.parseMegalodon(file, sample_name, mod_type = NULL, min_coverage = 5L)

Arguments

file

Character string. Path to the Megalodon per-read BED file.

sample_name

Character string. Sample name (used in messages).

mod_type

Character string or NULL. Modification type to assign to all sites (e.g., "6mA"). Megalodon files are modification- type-specific, so the type cannot be auto-detected from the file alone. If NULL, defaults to "6mA" with a warning.

min_coverage

Integer. Minimum read depth. Default 5.

Value

A data.frame with columns: chrom, position (1-based), strand, mod_type, beta, coverage.

Details

Megalodon per-read output (modified_bases.5mC.bed or similar) has the format:


  chrom  start  end  read_id  score  strand  ...  mod_prob

where mod_prob is the per-read probability of modification. This function aggregates across reads at each site by computing:

  • beta = mean of per-read probabilities at each site

  • coverage = number of reads overlapping each site

Sites with coverage < min_coverage are dropped.