about hg38

[FromSoSimple]

Thanks for sharing this tool. You mentioned we need to do install_github('aroneklund/copynumber') in order to run GRCh38. My guess is that this is to allow sequenza.extract function to take hg38 assembly because originally it only takes hg19, hg18, hg17.

Having checked your source code scar_score.R and preporcess.seqz.R, it seems that function sequenza.extract in preporcess.seqz.R does not take any parameter on assembly, so sequenza.extract would use hg19 (default) anyway no matter what reference (grch38 or grch37) you specify in scar_score.

If so, install_github('aroneklund/copynumber') is not needed to run on GRCh38. Am I understanding this correctly? Looking forward to your clarification.

Best, Corey

[squigzzz]

I have the same question Corey shared here, I was essentially wondering if the updated copy number package is only needed to run against GrCH38 or could we also use it to run against GrCh37 because I am getting drastically different results when I run scar_score using the modified copy number package vs. the original. There is even a difference in the stdout that mentions copy number segments whereas running it without the modified copy number package only outputs the variants, homo/het positions.