I'm currently using the ChIPQC package on a set of bam and peak files generated via bowtie2 and macs2 and then read into ChIPQC via a simple samples sheet. After running the program to examine some of the measures that it calculates, I am a little surprised by a few of the values, and I was hoping that somebody here might be able to explain them to me slightly, as google is not much help.
The following are the metrics from the troublesome .bam file:
Reads Map% Filt% Dup% ReadL FragL RelCC SSD RiP% D_1a 2011555 100 0 69.80 75 300 -181.00 1.350 1.050
I am mostly shocked by the measure here for RelCC and I have read that RelCC is calculated as follows:
RelCC = CrossCoverageScore(max) / CrossCoverageScore(readlength)
As I had understood it, neither of these values can drop below 0. I initially thought that the program was failing to properly calculate the metrics due to the very high level of duplication in this particular sample, but de-duplicating the .bam file and re-submitting leaves me with precisely the same RelCC score of -181.00.
Can anybody lend me a hand in figuring out if this is correct? And if it is correct, what does it mean? Did my lab manage to produce the literal worst ChIP-Seq of all time?
Thanks in advance for any help.