When the requisite genomic and phenotypic data are available for a large number of species, the next challenge is to identify the specific set of species that will help guide discovery of particular trait-relevant genomicregions. Broadly, the requirements are twofold: analogous to the structuring of GWAS, some of the genomes must be from species that do have a trait and others must be from species that do not have that trait. Second, both the group of spe cies with the trait and the group of species without the trait of interest should be as phylogenetically diverse as possible, to avoid conflation of genomic features that may be associated with general features that are shared due to common evolutionary history with genomic features that are essential to the trait of interest. With these criteria met, comparisons across mammalian species have proven sufficient to uncover the genomic basis of several phenotypes not detectable using museum specimens alone, including vision, as reported by Henrike Indrischek and colleagues in “Vision-Related Convergent Gene Losses Reveal SERPINE3’s Unknown Role in the Eye,” and the capacity for vitamin C synthesis, as revealed by Michael Hiller and colleagues in “A ‘Forward Genomics’ Approach Links Genotype to Phenotype Using Independent Phenotypic Losses among Related Species.” As the paper demonstrates, Vitamin C synthesis is a trait that is present in most mammals but whose loss in a few species, including humans, accounts for the tendency of people without access to fruits and vegetables to develop the well-known condition of scurvy.