 So, our work in tomato over the last several years stumbled upon a very interesting case of where domestication and breeding has led to negative interactions between beneficial mutations. And what we found is that there was a mutation that arose in the gene thousands of years ago that benefited one particular trait for the production of tomato, and this became widespread in the tomato breeding germplasm, as it's called, to the point where it was represented in more than 90% of all the breeding material that currently exists. Then about 70 years ago, a new mutation arose that had a benefit for a separate trait. And when breeders attempted to work with this new mutation that they discovered, they found that by crossing it into the other varieties, they had a dramatic negative interaction on productivity. Specifically what happened is that the plant started to overproduce flowers by making too many branches on the flowering shoots known as inflorescences. And normally this would be considered a beneficial trait, more flowers, more fruits coming from more branches. But in fact, this leads to an imbalance in the production of vegetative aspects of growth and reproductive aspects of growth. And this imbalance shifted towards more reproductive aspects of growth, flower production, is so extreme that the plant can't handle setting all of those flowers into fruits and leading to higher yield. So in fact, by breeding with the new beneficial mutation in the background of the old beneficial mutation, you actually had a reduction in yield. So when we decided to understand this negative interaction, this epistasis between these two beneficial mutations that together gave a negative effect on the plant, we were able to dissect the genetic interaction, identify the specific genes and mutations that were responsible, and essentially neutralize that negative interaction. Even more than that, once we identified the genes and identified the specific mutations, once we neutralized those negative interactions, we understand the genetic architecture of how these genes were working together to control flower production, and we were able to then exploit the system by mixing and matching the genes in precise ways through natural crosses. And this natural crossing and mixing of these mutations or combining of these mutations in different ways than breeders had done previously allowed us to create very weakly branched inflorescences. In other words, you're not causing as much of an imbalance on the plant between vegetative and reproductive growth. This weak branching then translates to a nice bump in flower production, and this bump in flower production then translates to higher yield.