So a lot of our atomic systems biology work has direct relevance to cancer, especially getting at mechanistic bases and understanding the role of siganlling and feedback circuits, and of course CANDO goes right at cancer related proteins and small molecules. However we don't think much about vaccines and while it is just coming to light, generalised vaccines have been the third pillar of my vision to improve human quality of life ever since Gödel, Escher, Bach (the first two being understanding basic mechanistic process at the atomic level, and therapeutic discovery using the basic science focussed on treatment and curing a preexisting condition). Only recently, in part due to Manish's efforts, has vaccine development come to the forefront and only as I spoke with Manish last week that we came up with a way to do computational science here, i.e., we can judge the effectiveness of a vaccine with our key hypotheses in a computer in a properly controlled manner (which of course still will need to be prospectively tested at the bench for maximum rigour).
In fact, from this, an epitope based RAPDF like function can be devised (which is really what Manish is doing informally) and THAT can be used to dri ve the simulations (somewhat circular) instead of vanilla RAPDF. I actually think for the first time that the vaccine design stuff can work, in terms of producing neutralising antibodies. If we can do this then we have both the drugs for after disease start and vaccines for prior to disease starting.
The epitope based vaccine design idea can be extended to target cancerous conditions, i.e., antibodies against cancer. This can be signalling molecules, whole cells (i.e., stuff that appears on cells during cancer), ... anything that makes something "cancerous" from a molecular phenotypic sense can be characterised and then an antibody vaccine designed against it using our approach.