 Another way to screen through molecules is to cheat, because in the computer I'm allowed to tear a molecule into pieces. So instead of trying to dock the entire molecule at once, I can take a molecule, this one that I want to test, and then I turn this into one, two or even three pieces, and then I start to check where will the first piece fit, where will the second piece fit, and where will the third piece fit. And then I see, is there a way that I can combine these pieces together? But there is no need for me to test this one molecule at a time, right? My goal here was to find something that fits. So maybe I can just try to test pieces separately. And then I will gradually, if I get a type, a piece F here that works well, and a piece Z in the middle that works well, and a piece, let's say K that works well. Well, that sounds like I should try the molecule FZK. That is likely going to bind. Now, that molecule doesn't exist in the lab. So this we will have to take to our organic chemist and try. But this so-called anchoring in a piece, and then gradually growing the molecule, that allows me to adapt every new piece of the molecule to the next cavity in the protein. Maybe I can fit an extra NH3 group here. Well, let's add in an NH3 group. So maybe FZKN would fit even better. Anchor and grow, build molecules in the receptor one by one. Just as before for docking, sloppy as our middle name. We should go for fast, fast, fast. If you think that it might be good, that's fine. Divine inspiration is okay here, as long as the probability is slightly better than zero of getting something.