From memory, organic molecules have rigid structures but also bonds that rotate so macro-molecules can changes shape to envelope a ligand. This is sometimes called the induced-fit model. My general conception of which is a bit like a catcher's mit.

Because of thermal energy equilibrium, a ligand is moving much faster relative to an enzyme. So it'll arrive at break-neck speed and smack into the catcher's mit and then probably bustle around a bit. The ligand and its associated water bubble (closely interacting water molecules) will interact with arms of the enzyme, attracting them to move and grasp the hydrated ligand. Eventually this movement will displace the water molecules, leaving only the chummy direct embrace occuring between the ligand and the protein. They are compatible and decide to make a night of it. So the big one is hugging the little (or the mit is holding the baseball). At this point my dualing metaphors break down because they interact chemically, often with the help of a phosphate molecule to power the little (nano) machines into doing some metabolic magic.

They then decide that they need some time apart to get their heads straight and find themselves, leaving only an ambivalent ADP molecule looking on as they make their separate ways in the endless hustle and bustle (thermal agitation) of the molecular realm.

My metaphor of the catcher's mit seemed to metaphorphose into more of a romantic interlude but you get the idea. Molecules in this situation are basically like ravers in a mosh-pit who are attracted to one another, hug, make out briefly, then realise that they have to find their friends.