Predicting the relationship between protein sequence and function remains a major challenge in genomics. Homologs that are divergent in sequence often have conserved functions, while rare single nucleotide polymorphisms in a population are the basis of many human diseases. Variation in protein-coding sequence can impact function in various ways, such as by altering folding, enzyme kinetics, binding affinity, or other factors. Here I present a method I am developing to investigate the effects of sequence divergence on the functions of proteins from both retrospective and prospective viewpoints. This method will enable the characterization of full-length protein variants expressed from the native genomic locus through quantification of variant frequencies in response to competitive growth in selective conditions. Here, using Saccharomyces as a model genus, I show how this method will illuminate the evolutionary fitness landscape of a small protein (Atx1) from all Saccharomyces species and homologs from other yeasts and multicellular eukaryotes, including humans. Using established genome-editing technologies, I will generate strain libraries that are completely isogenic, except at the ATX1 locus. I will then conduct pooled competitions of cells expressing ATX1 variants in selective conditions and monitor the changes in variant frequencies by high-throughput sequencing. Deriving fitness values from these frequency changes enables the comparison of the effect of a given genetic variant on protein function and, ultimately, illuminates the constraints present in the fitness landscape of ATX1. Thus far, measuring growth of yeast strains expressing individual Atx1 homologs in selective conditions has revealed differences in fitness, including a general trend toward greater sequence divergence resulting in lower fitness. Notably, HAH1 (the human homolog of ATX1) is able to complement the yeast deletion. Pooled competitive growth assays will further reveal the fitness consequences of genetic variation as cells expressing Atx1 homologs compete for limited resources. Additionally, testing rare ATX1 variants generated by saturation mutagenesis will ultimately enable the characterization of a fitness landscape for ATX1. This technology could be used to study genes implicated in human disease and serve as a platform for diagnosis of heritable diseases that are due to hypomorphic alleles that are present at low frequencies in humans. With the emergence of precision medicine enabled by affordable genome sequencing, physicians could use data generated by this type of study to diagnose a wide variety of genetic diseases using only the patient’s genome sequence.