Alan Boyle

Yes, this makes sense - I will compute these and share them in the repo.

The smaller genomes use a different overlap parameter. I recommend using 5kb instead of 20kb like was done for human. You can change this in the code here: https://github.com/Boyle-Lab/Blacklist/blob/master/blacklist.cpp#L469

Have you been able to work out the discrepancy? Perhaps we can compare md5 hashs of the bam files to make sure that we are working with the same data?

10.2.1 Here is my env: ``` # packages in environment at /home/apboyle/.conda/envs/b529: # # Name Version Build Channel _libgcc_mutex 0.1 conda_forge conda-forge _openmp_mutex 4.5 2_kmp_llvm conda-forge anyio 4.2.0 pyhd8ed1ab_0 conda-forge...

Can you provide a link to the bed file? One thought is that there is an aberrant read position throwing things off because it is much larger than the chromosome....

To avoid this error you will need to use an older version of Java.

Since this software is only intended to be run once per organism, we haven't felt the need to install it on HPC infrastructure. Instead, you most users simply download the...

I don't think that you will get a functional blacklist with this tool from only 12 input bams, so it is not worth generating that. As far as bias introduced,...

I think you probably need more than 100 input tracks for a quality blacklist. We used 1256, 287, 443, and 490 for human, mouse, fly, and worm.

It may be that there is very low signal in this chromosome? If you uncomment like 442 you can see what the thresholds are defined as and that might give...