This application claims processes and compositions that enable discovery of single nucleotide polymorphisms (SNPs) and other sequence variation that follows two essentially identical sequences, one a reference, the other a target, as well as SNPs discovered using these processes and compositions. The inventive process comprises preparation of four sets of primers, "T-extendable", "A-extendable", "C-extendable", and "G-extendable". These primers, when templated on a reference genome, add (respectively) T, A, C, and G to their 3'-ends. The invention also comprises a step where these primer sets are separately bound to complementary sequences on target DNA and, once bound, prime extension reactions using target DNA as the template. If the target DNA directs incorporation of the same nucleotide as the reference DNA, then the T-, A-, C-, and G-extendable primers are extended (respectively) by T, A, C, and G. The architecture of the process distinguishes products from these extensions from products derived if not T, not A, not C and not G ("3N" or "3", to indicate the other three nucleotides) are not added. Thus, this process discovers differences between the target and reference DNA in the site queried by the primer extension reaction. The distinction makes the two kinds of products either separable or differentially extendable. This distinction is used to disregard products that added T, A, C, and G and to identify the sequence(s) of primers that added not-T, not-A, not-C, and not-G. Further and optionally, information from these sequences identifies loci of the SNPs in an in silico genome.