bigwig-signal-processing

bigwig-signal-processing

Summary

Convert bias-corrected ATAC-seq cutsite signal into per-base footprint scores by measuring insertion depletion patterns within accessible chromatin regions using bigWig format. This skill quantifies transcription factor occupancy by computing the magnitude of Tn5 insertion depletion at each genomic position.

When to use

Apply this skill after bias correction of ATAC-seq reads (via ATACorrect) when you have a bias-corrected bigWig file and need to measure transcription factor footprint strength within defined accessible regions (peaks, motif sites, or called footprint boundaries). Use it as the intermediate step between bias-corrected signal and downstream binding site detection or visualization.

When NOT to use

• Input bigWig file has not been bias-corrected for Tn5 transposase sequence preference—raw or only-quantile-normalized ATAC-seq signal will produce misleading footprint scores.
• No defined set of accessible regions is available; ScoreBigwig requires a BED file of boundaries and cannot discover footprints de novo.
• The analysis goal is only visualization or aggregate signal inspection—simpler tools like PlotAggregate may be more direct without computing per-base scores.

Inputs

• bias-corrected bigWig file (*.bw)
• accessible chromatin regions in BED format (peaks, motifs, or footprint regions)

Outputs

• footprint score bigWig file (*.bw) with per-base depletion magnitude scores

How to apply

Load the bias-corrected bigWig file (output from ATACorrect, typically named *_corrected.bw) and provide a set of accessible genomic regions in BED format (ATAC-seq peaks, motif regions, or previously defined footprint boundaries). Use TOBIAS ScoreBigwig to compute footprint scores by measuring signal depletion—the tool aggregates the corrected insertion signal within each region and outputs a new bigWig file where each position's value reflects the magnitude of the local depletion pattern characteristic of protein-DNA binding. Validate the output bigWig for correct format (proper header, coordinate system consistency) and non-empty score distributions across regions; scores should show characteristic troughs at true binding sites and elevated signal in flanking regions.