Devon Ryan

Can you post the SAM entries for those 6 reads?

Ah, interesting. I wonder if this is due to the reverse strand reads starting before the forward strand reads in those two cases. Dovetailing is biological impossibility in Illumina sequencing...

Can you let me know what version of pysam you have installed? The behavior I expect to see in the code is that only the overlapping portion of the reads...

These lines should suffice. There is a docker container from the biocontainer project for each version of deepTools (they're auto-generated by bioconda).

OK, I've found the cause. Internally we do some sanity checking rather than blindly believing the flags set by the aligner. In this case the dovetailing internally changes the "proper...

Samtools and pysam simply report what flags are set, they don't do any sanity checking. Usually it's the 3' end of reads that are a bit wonky, not the 5'...

These are regions for where this is no data in your bigWig file(s). Depending on what sort of data they contain, you may want to set those values to 0...

You might be able to run `nosetests --with-doctest -sv deeptools`, presuming you have `nose` installed. That's done as part of our automated testing and generates some plots that you won't...

Is there any reason your scores span such an extreme range? It's column 4 of you BED file that's causing an issue.

Well you will end up running into weird corner cases like this due to treating it as a 32-bit integer.