ABSTRACT
This chapter introduces the notion of reduced-order modeling and demonstrates its applicability in the estimation of antibody affinity. It is known that antibodies produced during the immune response are heterogeneous with respect to their affinity toward the antigen. The chapter reviews the problem of determining the affinity distribution into an equivalent problem of identifying a parametric model of a linear system. It develops a numerical procedure to carry out the model identification along with computer-simulation results. The chapter then presents the concept of "system inverse". It discusses some possible topics for further estimation of antibody affinity in the area of reduced-order modeling. Depending upon the numerical method used, the affinity density may be unimodal or bimodal. Some preliminary results have been presented to highlight the applicability of reduced-order modeling concepts known in linear system theory to the determination of antibody affinity.
