Gene-specific selection with counting method when no synonymous subsitutions

Question

[ksw]

Hi, I am using short-read data to look at pathogen deep-sequencing data and look for signatures of selection within this pooled data. I am using the counting method to estimate dN/dS ratios on specific genes within the population of samples (I do not have haplotype information).

I was wondering how to calculate dN/dS ratios if there are 0 synonymous substitutions at a particular locus of interest?

Thank you!