Antibiotic resistance.
Here's my roadmap: Solve gene sequencing (done). Solve protein folding. We just "solved" that, though for what I have in mind, we need more. We need to be able to, from a bacteria's genome, determine enough of the expressed proteins to find what that bacteria is vulnerable to.
Then, last step: Given a bacteria with a particular DNA, from the proteins, find an antibiotic that will kill it.
My vision is for doctors' offices to be able to take a sample, somewhere local be able to do the DNA sequencing, and if there's no known effective antibiotic for it, send the sequence data to a supercomputer at CDC that solves the protein folding problem, and from there derives a new antibiotic that will be effective at killing that bacteria. I'd like the whole process to take no more than hours. Even better if there's somewhere local to the patient that can take the CDC's output and synthesize some of the new antibiotic. (And, of course, the new result goes into the library of known solutions to known bacteria.)
