Parabon NanoLabs has done some research that's a bit different from other face-shape studies. For example, they've done big-data-type analyses on 1 million distinctive letters in the human genome, called single nucleotide polymorphisms, or SNPs. They searched for single SNPs, as well as combinations of up to five SNPs, that are associated with different face shapes. Many studies seeking SNPs that affect physical traits look at only single SNPs or pairs of SNPs, but it's thought that most human traits are probably controlled by numerous genes, so Parabon's methods could have a better shot at meaningful associations than similar, past studies. There's a big limitation, however. Computers simply can't test every possible combination; that's too many calculations. For example, there are 170 quadrillion possible three-SNP combinations in a set of 1 million SNPs. So Parabon uses an algorithm to choose some smaller number of combinations, which it believes are most likely to be important, to test.