I would love to build systems that modeled pathogens and applied monte carlo methods/fuzzing techniques to permuting proteins or other molecules that might render the pathogens harmless within the simulation. One might then use these results to either cure the pathogen else refine the model. Sort of like TDD for biology. I have no idea where to begin with modeling such profoundly complex biological systems, but please let me know if you have any suggestions because I would love to pursue this.
Things like this are being done, generally under the umbrella of "Computational Biology" or occasionally "Bioinformatics" (a field which is still sorting out what it is).
Doing it well is going to require a great deal of education.
