Hi, I want to normalise the allele-specific read counts across the samples and the goal is to see the differences in the expression of say ref.allele across all the samples, or in the alt.allele across all the samples.

What is needed is:

ref. allele1 vs ref.allele2 vs ref.allele3 alt.allele1 vs alt.allele2 vs alt.allele3

Can you please share more details on your experimental setup? Also, how does this relate to DESeq2?