ABSTRACT
Using idiotypic networks as an example, I propose an approach to developing realistic models of the immune system. I show that systems containing large numbers of different clones, with particular clones being created and destroyed, can be studied by means of computer simulation. By labeling B cell clones and the antibodies that they secrete by a binary string, and using rules of complementarity between binary numbers, I show that one can create complex models with realistic topologies. I point out that the language of graph theory may be useful in characterizing idiotypic networks. I also give an example of a dynamical system of equations, involving antibody, B cell clones and growing antigen that incorporates many of the features of T-independent immune responses. To fully specify these dynamical equations, many important problems in the chemistry of multicomponent mixtures of anti-idiotypic antibodies, B cell receptors and antigen need to be solved. By viewing the immune system as a dynamic entity, whose components may continually be created and destroyed, a new view of the immune system emerges. I suggest that under some circumstances, the immune system may not be operating at a steady-state which is then perturbed by the presence of antigen, but rather may be constantly changing. Thus, like a weather pattern, the immune system may be rather quiescent 378at some times but raging at other times. Lastly, I suggest that memory can not only be stored by memory cells, but may also be stored dynamically in idiotypic networks.
