I have a bed file with genes I am interested in and Bigwig files from my ChIP-Seq experiment after subtracting input from treatment files using deeptools bamcompare. I used the bigWigAverageOverBed utility from UCSC to calculate the coverage. I only want to keep those that have a good mappability. What might be an appropriate cut-off for the scores from bigWigAverageOverBed ? The scores range from -7030 till 71800.
Also, I want to keep all negative values. So I cannot perform a log transformation. Is there any other way to scale or normalize the data so that I wont lose relevant information ?