How can I use MIDAS to get pN/pS on a per-gene basis?

[JuliaMcGonigle]

Per the snp_diversity.py script, documentation states: Only quantify diversity at synonymous sites (Note: use the ratio of 5/6 for pN/pS)

What does this mean? Is pN/pS calculated by running: snp_diversity.py /path/to/snps --genomic_type per-gene --site_type 1D --locus_type CDS --out /path/to/output

and then: snp_diversity.py /path/to/snps --genomic_type per-gene --site_type 4D --locus_type CDS --out /path/to/output

And then taking the pi value in the output for 1D/pi value in the output for 4D?

What does the ratio of 5/6 have to do with it?