Whole Genome Variant Calling on Partially Masked Genomic Data

[sdgagephd]

Hello,
I have been using your Whole Genome Analysis Pipeline to generate GVCF- and VCF-formatted data from previously mapped CRAM data. I am looking to also use it for partially masked, whole genome CRAM datasets, which make up ~10% of my cohort. Unfortunately, the masking has been making the alignment step, required for unmapped BAMs, crash the workflow. I have gotten a couple of suggestions for troubleshoot; the most practical seems to be to use the masked BAMs before the unmapping step. Since my data is already mapped, is there a way I can pass the mapped, masked data into your pipeline? Thank you!
Steve