 So to understand these structural features and why evolution happens we need to understand why some folds are stable And why some folds in particular are more stable than other folds. I'm going to need to show you an example fold here It's a Rossman domain, but it doesn't matter in this case What things are good and tend to make fold stable? Well, there are a couple of things First you need many hydrogen bonds in particular the saturation and the helices that all all amino acid residues Participate in hydrogen bonds. That's awesome. In particular in the interior of the protein if we had unpaired hydrogen bonds That would be very unfavorable They either need to bind to water molecules or they need to bind to something else that is hydrogen bond partner We don't want these long loops to exist in the interior partly for their reason of hydrogen bonds But also because they're not going to pack very well and they would also the loops are kind of floppy, right? So that that means that there's lots of entropy there if you take those loops and put them interior We would lose a lot of entropy bad in terms of free energy We also want the edge of the helices and sheets to face water That's partly related to this hydrogen bond pairing But again when we are at the edge of a sheet, we're no longer forming more hydrogen bonds. It's much better to let that face water Helix and sheet regions should usually be a bit separate and by separate I don't mean separate as in one alpha part and one beta part here But so we actually have one complete beta sheet It happens to be stuck between kind of two layers of alpha helices But there is a little bit this joint and that's because we can pack 20 residues in an alpha helix And I can or I can pack say 10 or 15 of them in a sheet But the more I mix them the more complicated it gets so somewhat segregated and Then I need to make sure that I don't have too many distortions So putting a proline in a helix technically I can do it but then I would get that kink in the helix I might be able to survive one such defect, but as you will see in a few slides Each such defect is costly and having more than one is virtually impossible