Predicting plant-specific small molecule binding variability across species

Agricultural plants must resist myriad stresses such as heat, disease, drought, and herbivory. Plant biochemists are highly incentivized to discover chemical methods of improvement that will mitigate these stresses. Finding novel methods of improving plant function and resistance is difficult and time consuming, however, by taking advantage of millions of years of evolutionary adaptation, we may be able to accelerate research. Here, we use Alphafold, NAMD, Deepsurf, and Autodock Vina, to construct a computational pipeline that that uses amino acid sequences and ligand structures to rapidly screen large panels of small compounds such as isoprene, ethylene, and auxin, for potential binding sites as well as binding affinity on plant proteins. From these binding affinities, we can determine likely targets of small compounds, and whether compounds currently in heavy usage may have off-target effects that have not previously been investigated.