Biparatopic antibodies are bispecific antibodies (BsAbs) that bind distinct, non-overlapping epitopes on the same antigen. Their increased binding affinity leads to an enhanced anti-tumor effect, such as increased immune cell effector function, increased lysosomal trafficking, and downregulated proliferative signaling. One form of a biparatopic antibody is a radioisotope-engaging BsAb, which pairs single-chain variable fragments (scFvs) for specificity with a radioisotope chelator for efficacy. 

There are several computational tools that inform antibody design. Alphafold is a neural network-based model to predict 3D protein structure based on amino acid sequence. Meanwhile, Molecular Mechanics - Generalized Born Surface Area (MM-GBSA) calculates the Gibbs free energy change (dG) of antibody-antigen binding by summing thermodynamic interactions (covalent, electrostatic, polar, and van der Waals forces) across all atoms. Alphafold is used to map antibody-bound antigen epitopes, while MM-GBSA predict binding affinity of the antibody-antigen complex.

My methods emulated a funnel, which started with a large number of candidates and progressively screened them through increasingly specific criteria. First, I took a library of pre-existing antibodies and characterized their properties. Next, I would then input pairs of antibodies into AlphaFold to map their binding epitopes. If the binding sites were compatible and the interdomain spacing was good, I would run MM-GBSA calculations. If the MM-GBSA calculations indicated high binding affinity, the antibody would be a viable candidate for downstream testing. 

I performed several rounds of optimization to generate strong candidates. For proper spacing, I adjusted linker length or added an extra radioisotope chelator separating the two scFvs to ensure they bound their correct epitope. The extra radioisotope chelator is also promising because it increases the payload delivered to the tumor. Overall, this yielded high-specificity, high-affinity antibodies that are promising candidates for immunotherapy.