Filling the Gaps in the Chip-seq data

Question

[eneskemalergin]

Hello all,

I have a chip-seq data and I realized that there are some gaps. It looks like its going with span 25 but in some parts there is no score available. Why is that?

If I want to fill those gaps with -let's say- 0, how can I do that? ()

This is the sample from the data

chr21   9411475 9411500 1
chr21   9411500 9411525 1
chr21   9411525 9411550 0.96
chr21   9413025 9413050 0.64
chr21   9413050 9413075 1
chr21   9413075 9413100 1
chr21   9413100 9413125 1
chr21   9413125 9413150 1
chr21   9413150 9413175 1
chr21   9413175 9413200 1
chr21   9413200 9413225 1.36
chr21   9413225 9413250 1.36
chr21   9413250 9413275 1
chr21   9413275 9413300 1
chr21   9413300 9413325 1
chr21   9413325 9413350 1
chr21   9413350 9413375 2.44
chr21   9413375 9413400 3
chr21   9413400 9413425 2.64
chr21   9413425 9413450 2
chr21   9413450 9413475 2
chr21   9413475 9413500 2
chr21   9413500 9413525 2
chr21   9413525 9413550 2
chr21   9413550 9413575 0.56    <- from here it skips 400 bp
chr21   9413975 9414000 1.32
chr21   9414000 9414025 2.8

Thanks...

Answer 1

[dariober]

I think bedtools complement comes in handy here to extract the gaps. Pipe the gaps to awk to add 0 as coverage, then merge sort the peaks and gaps in a single file (not tested):

bedtools complement -i peaks.bed -g chromSize.genome \
| awk '{print $0 "\t0"}' > gaps.bed

sort -k1,1 -k2,2n peaks.bed gaps.bed > peaks_gaps.bed && 
rm gaps.bed

Answer 2

[Carlo Yague]

The gaps probably come from regions with zero coverage, but its hard to tell with certitude without knowing where your data comes from.

I don't know an easy way to fill the gaps, although writing a little script to do that should not be too hard. [EDIT : see dariober's answer :) ]

An alternative could be, if you have access to the original alignment files, to use bedtools genomecov with '-bga' option to rebuild the bedgraph file with the zeros.