Vaccines have been valuable tools in the prevention of infectious diseases, and the rapid development of new vectors against constantly mutating foreign antigens in viruses such as influenza has become a regular, seasonal exercise. Harnessing the immune response against self-antigens is not necessarily analogous or as achievable by iterative processes, and since the desired outcome includes leaving the targeted organism intact, requires some precision engineering. In vaccine-based treatment of autoimmunity and cancer, the proper selection of antigens and generation of the desired antigen-specific therapeutic immunity has been challenging. Both cases involve a threshold of existing, undesired immunity that must be overcome, and despite considerable academic and industry efforts, this challenge has proven to be largely refractory to vaccine approaches leveraging enhanced vectors, adjuvants, and administration strategies. There are in silico approaches in development for predicting the immunogenicity of self-antigen epitopes, which are being validated slowly. One simple approach showing promise is the functional screening of self-antigen epitopes for selective Th1 antitumor immunogenicity, or inversely, selective Th2 immunogenicity for treatment of autoimmune inflammation. The approach reveals the importance of confirming both Th1 and Th2 components of a vaccine immunogen; the two can confound one another if not parsed but may be used individually to modulate antigen-specific inflammation in autoimmune disease or cancer.