A buddy of mine was working in the lab of the scientist who "invented" unnatural amino acid technology (rumblings about how it was taken from the postdoc in the lab across the hall). That was the sexy project de jour of the lab; but my buddy was working on a less sexy project - catalytic antibodies. This is where you take the concept that antibodies can be adapted to fit any molecular shape and try to make it fit something that looks like the transition state of a chemical reaction, resulting in acceleration of the chemical reaction. (designer enzymes!)

The results till that day were modest at best, and that's because stabilizing the transition state is only part of what a real enzyme should do, and even though the professors are supposed to teach you this in your biochem/chemical biology deep-dive courses in grad school YMMV, and it's easy to sell a starry eyed grad student, especially when the prof doesn't know any better, too.

Anyways my buddy's project was even worse - he was supposed to make a protease (an enzyme that degrages proteins). And if you look at proteases, their clefts wrap around the protein even more than an antibody ever could (they have shallow clefts), because burying the reaction away from water is a critical aspect of their function.

He spent three years working in a lab that demanded 80+ hour workweeks. Towards the end his sleep cycle had flipped, he was playing around making geometric designs with his pipet tips, and spending much of his workday playing a flash website gameboy tetris, and many days going to the casino to play poker instead of work.

finally his boss modified his project, instructing him to graft a metalloprotease domain, onto the antibody in an attempt to get it working. A breath of fresh air! Suddenly he was invigorated with a new approach to the project. But not long after that, he was back to the old routine of being burned out, and totally unproductive, spending hours on trivialities, like trying to strip metals from his water supply to really get it right and get it working. In the end, it never worked.

It turned out that the metalloprotease domain was designed by Homme Hellinga. Years after this, the scientific community discovered that Homme Hellinga was faking his enzyme design work.