Specialized metabolite genes can arise by gene duplication and neofunctionalization where one copy retains the ancestral function while the other changes to a new function. This process is at the core of chemical diversity across species. These duplicates also enable variation within species often from presence-absence-variation (PAV). The combination of PAV in reference genomes and the potential for convergent evolution could greatly complicate the use of sequence homology to identify enzymes across a plant family. To explore this potential, I use the GSL-OH enzyme in the glucosinolate biosynthetic pathway as a case study. Glucosinolates are a class of specialized metabolites with the most diverse representation in the order Brassicales. They play roles in plant environment interactions and humans perceive them as the flavor compounds in broccoli, wasabi, mustard, amongst other vegetables. GSL-OH catalyzes a terminal structural modification reaction that converts the AOP2 created but-3-enyl into S-2-hydroxy-but-3-enyl (S2HB3) and R-2-hydroxy-but-3-enyl (R2HB3). To identify GSL-OH in non-model species I first identified potential homology based GSL-OH candidate genes across the Brassicacea family. To test for potential convergent evolution where a non-homologous enzyme evolved the GSL-OH reaction, I used a transcriptomics genetic covariation approach. This consists of using transcriptomics in Brassica rapa and Brassica oleracea accessions with PAV for 2-hydroxy-but-3-enyl presence. Given the potential for the reference genome to be missing the GSL-OH gene we are developing a kmer approach to identify genes that correlate with the 2HB3 metabolite data for these Brassica accessions. The candidate genes from both approaches have been cloned and are being used to test for complementation in Arabidopsis thaliana lines that are natural knockouts for GSL-OH. If the transgenic lines produce 2-hydroxy-but-3-enyl, we can conclude that the gene is responsible for the GSL-OH activity. Thus far I have identified two previously uncharacterized GSL-OH genes from the plant Descurainia pinnata that have evolved novel sterochemistries. One produces a 5R:1S ratio of the enantiomers while the other enzyme results in a 1R:5S enantiomer ratio. Analyzing Descurainia plant tissue, showed a 4:1 ratio of R: S suggesting that the enzymes functions may both function and unequally contribute to the pool of 2-hydroxy-but-3-enyl enantiomers. Moving forward using the method I have developed I will continue to characterize candidate genes for the glucosinolate pathway to help inform our construction of phylogenetic tree for genes of this metabolic pathway as a way to explore the evolutionary patterns that have shaped this branch of specialized metabolism.