For a few cases maybe, but probably not for most. Ka is the number of non-synonymous base changes and Ks is the number of synonymous base changes in the same gene. In most cases of within species gene polymorphism no synonymous base changes present (say Ka=1 and Ks= 0, then ratio is 1/0, which cannot be).

I don’t know if you are still checking this board but you can use Ka/Ks for intraspecies comparisons. Here is the paper that developed the idea. Hughes AL, Nei M. 1988. Pattern of nucleotide substitution at major histocompatibility complex class I loci reveals overdominant selection. Nature 335:167-170.

The software, KaKs_Calculator, can do this. It contains several existing methods and can be found at http://www.yale.edu/townsend/software.html