I am under the impression that typical NGS workflow does not allow us to identify zygosity of somatic deletion ( I recall something called phased sequencing can achieve this by tag strandness of reads ). Is there any algorithic solution for such task?
What does zygosity mean for a somatic variant? Do you want to know the zygosity of the variant, just for those cells affected by it?
Yes. Whether the variant occurs on the same allele of these cancer cells.