Recommended `--exclude` BED in the detect step

[SHuang-Broad]

Hi,

I am wondering if you have any recommendations on what should go into the BED that is fed to --exclude in the detect step. I am thinking about segmental duplications.

Thanks! Steve

[paudano]

I don't use this option. Instead, I run SMRT-SV with no filtering, but I'll filter the SV calls afterward. That way, I can always go back and look at what was removed, which cannot be done if it filters in SMRT-SV. If you filter segmental duplications, then you are losing too many valuable loci that long reads are capable of resolving.

I created a "low-confidence" filter on GRCh38 to remove SV calls in centromeres and peri-centromeric loci where SMRT-SV reproducibility was low (10.1016/j.cell.2018.12.019).

Filter BED: lc.bed.gz