MACS2 Narrow peak (narrowPeak) output format

Question

[ariel.balter]

According to the macs2 documentation, column 7 of the *.narrowPeak output bedfiles is supposed to be the fold change. However, using bedtools multicov I get a different result.

$ bedtools multicov -bams ../bam/mock_lsd1_A.bam ../bam/mock_input_A.bam -bed mock_lsd1_A_peaks.narrowPeak > mockAcoverage.bed
$ awk 'OFS="\t"{print $7,$11/$12}' mockAcoverage.bed | head -20
5.29438 5.22973
7.67026 5.22973
3.65672 6
9.81878 10.2857
4.15265 5.05
2.92799 2.625
3.21866 6.16667
3.29258 6
7.35857 8.28571
4.37685 8.28571
4.24169 8.69231
6.47838 8.69231
10.42606        8.69231
2.63428 3.1
3.56365 6.66667
3.51999 6.25
3.88036 5.625
4.59437 5.625
6.13598 5.46154
2.99039 4.16667

Another post on a similar topic.

[ejm32]

I am pretty sure the reason the fold changes are different is because in macs2 the denominator is a function of lambda and not a simple division of IP/input for a given peak region. From the macs2 github page (while this is for the xls, one would have to assume that it applies to the *.narrowPeak file as well):

fold enrichment for this peak summit against random Poisson distribution with local lambda, -log10(qvalue) at peak summit