Figure 1.

T-cell counts corresponding to three different but similar simulations. In panel (A) we inject two homologous viruses (i.e., cross-reacting) in succession (i.e., at day t_I = 0 and t_II = 330) namely ${V^I=V}_0$, $V^{II}=V_2$, which are therefore at a distance of 2 bits. In panel (B) we inject two heterologous viruses, V₀ and V₈. In panel (C) instead, after the usual priming with V₀ we challenge the system with both the homologous V₂ and the heterologous V₈ viruses. Cell counts shown in all panels are representative of the most active clones of the 2^N possible. These “responding clones” are specific either to just V₀ (filled triangles in all panels A–C), or specific to both V₀ and V₂ (i.e., cross-reacting memory clones shown in panels A,C as empty squares), or naïve independent responses specific to V₈ (panels B,C, filled squares). In panel A the primary naïve response to V₀ (filled triangles) leaves memory that is then re-stimulated by V₂. In panel B instead, the same primary naïve response to V₀ (filled triangles) do not cross-react to V₈ and indeed fade away upon the raising of the secondary responses. (C) shows what happen when the system is challenged with both. It shows in particular that the heterologous response to V₈ (filled squares) is independent and unaffected by the cross-reacting clones (empty squares) operating the MaN.