 So let me show you a real chevron plot instead of my illustration first. They would look roughly like this. Again, we have three important regions. We have the region here on the left, where unfolding dominates. We have the region, sorry, where folding dominates. The region on the right where unfolding dominates. And we have the midpoint, whereas we measure these shifts on the curves. The classical example here is one called Barnais and Barstar, that Alan first spent almost entire career studying. And the idea is that we take a small protein and we systematically mutate one residue at a time. And then we get the new chevron curve and then we see how that changed these two slopes. Now, if you do this for Barnais and Barstar, you get a bunch of dots here. And these dots then define for that particular residue whether part of the transition state or not. So I say this is actually bacterial ribonuclease. That's why it's called Barnais, but it's not important for now. Based on that, Alan and colleagues have been able to identify the entire transition state, characterize it, build models. They've been able to confirm those models with collaborators doing molecular simulations lately. So we know everything about how the actual folding process happens almost, despite never having seen any actual structure of the transition state itself, because we can't determine that. That's pretty cool. Here's another example. Beta and outer membrane protein. So when we talked about membrane proteins, I kind of glossed over that there are actually beta sheet membrane proteins, but they're pretty rare. So in the outer membranes of some bacteria, you have membrane proteins that is just a barrel of a continuous beta sheet. But we know much less about how they fold. So maybe we can use this type of method for it. So when we have the structure as you see here, we start introducing mutations and then we determine a number of chevron plots as you see there on the lower right. Based on that, you would have some residues here that are definitely part of the transition state. And do you see how they tend to group in the middle layer here of the beta sheet? So this appears to suggest that there is a small component of the central part of the beta sheet that forms pretty much all the way around this porin. And then they gradually extend to both sides, which is likely how they are then inserted in the membrane. Again, this is based on models derived from the transition state. We can't determine the transition state structure itself. Last example from Mikael O'Liverberg, colleague at Stockholm University. They have been studying a number of proteins involved in neurogenerative diseases. This particular small protein is interesting because the N and the C terminus are very close to each other. So it's almost that if you just created a small link there, you could link the chain to itself. So it was infinite. They couldn't do that, that wouldn't fold. But they then ask, how important is it exactly where the N and C terminus break is? So maybe if we do link the N and C terminus together, but instead create an opening elsewhere. So that is, you're essentially doing a circular permutation of the protein. In that case, it turns out that the transition state is roughly the same. You have to first approximation the same residues involved. It folds in the same way, meaning the actual sequence in the chain is not important. It's the global properties of the entire protein. Pretty important.