With the advent of the era of precision medicine, monoclonal antibody drugs have become a hotspot in biological drug treatment due to their high specificity and effectiveness. As the types of monoclonal antibody drugs increase and their applications widen, related immunogenicity issues gradually emerge. For patients, immunogenicity affects the safety and effectiveness of drugs, and even brings fatal new diseases to patients due to ADA and endogenous protein cross-reactivity; for enterprises, the risk of research and development greatly increases, and if ADA problems are discovered in the late stage of clinical development, it will result in heavy losses; for drug regulatory departments, immunogenicity has also become a top priority, and all biological drugs must have an immunogenicity evaluation before they go on the market to ensure the safety and effectiveness of the drugs.

At present, although chimeric antibodies, humanized antibodies and fully human monoclonal antibodies have been developed in order to reduce immunogenicity, humanization has not completely eliminated ADA. Even the CDR region of humanized antibodies still has strong immunogenicity and produces unique type antibodies. We have developed an algorithm to predict antibody immunogenicity, which predicts whether the antibody peptide sequence is associated with human MHC2, whether the MHC2-peptide complex binds to TCR, whether the antibody peptide is associated with human B cells, etc., to judge whether the antibody has a risk of immunogenicity. If there is a certain risk of immunogenicity, we can also remove the immunogenicity of the antibody by designing point mutations at specific sites that cause immunogenicity. The specific process is as follows: