Er, if the recent decoding of the human genome has pointed anything out, it’s trashed the concept of “One Protein, One Receptor”. DNA is a ton more complex than has been taught for 30+ years.
Quantum computers have been at “just twenty more years” for at least the past decade. At this rate, we’ll have them just in time to be powered by practical fusion reactors.
If we knew what the next thing was going to be, then we would already have it.
Molecular computing is a more attainable goal (although only slightly more attainable); where each transistor in an IC would be a molecule, each 1 and 0 of computer memory would be represented by the presence/absence of a single electron etc.
Or biological computing, where, instead of a program, you feed the string of instructions in the form of a protein; various enzymes etc work on this in whatever way is natural to them (natural in the sense of that being what they were engineered to do) and the end result (also a protein) is your coded result. Remotely possible, I suppose, but don’t expect to play Quake on such a computer.