Figure 1.

Challenges of comparing antibody-virus datasets

(A) We develop a framework that predicts antibody responses (e.g., binding, hemagglutination inhibition [HAI], or neutralization) of any serum against viral variants from any other dataset, enabling direct cross-study comparison.

(B) Because each serum is unique and virus panels often only partially overlap, these expanded measurements are necessary to characterize the limits of the antibody response or quantify tradeoffs between key features, such as potency (the strength of a response) vs. breadth (how many viruses are inhibited).