bedgraph-file-format-manipulation

BEDGRAPH file format manipulation

Summary

BEDGRAPH is a tab-delimited format for storing continuous genomic signal (coverage, p-values, q-values) across regions, with one score per base pair. This skill covers generating, transforming, and combining BEDGRAPH tracks through sequential operations (pileup, normalization, comparison, merging) to produce score landscapes suitable for peak calling.

When to use

You have aligned ChIP-Seq reads (in BED or BEDPE format) and need to convert them into quantitative genome-wide signal tracks (coverage, p-value, or q-value scores) for downstream statistical comparison or peak detection. Specifically, apply this skill when you must: (1) generate pileup coverage tracks from reads extended to a known or predicted fragment length, (2) normalize background tracks to account for sequencing depth or local bias, (3) compute statistical scores (p-value or q-value) by comparing ChIP and control signal tracks, or (4) merge multiple background layers (fragment-length, 1kb local, 10kb local, genome-wide) into a single background estimate.

When NOT to use

• Input reads are already in BEDGRAPH or BigWig format — skip pileup generation and proceed directly to normalization or comparison.
• You are performing broad peak calling — use bdgbroadcall instead of the narrow peak pathway; background layer construction differs.
• Fragment length is unknown and unpredictable (e.g., highly variable insert-size library) — prediction may fail; consider external QC or alternative peak callers.

Inputs

• Filtered ChIP-Seq reads in BED format (from macs3 filterdup)
• Filtered control reads in BED format (from macs3 filterdup)
• Predicted fragment length d (from macs3 predictd)
• Genome size (integer, bp)

Outputs

• ChIP pileup BEDGRAPH (coverage track, one score per base pair)
• Control pileup BEDGRAPH files (d, slocal, llocal)
• Combined local lambda (background bias) BEDGRAPH
• Score BEDGRAPH (p-value or q-value track, one score per base pair)