Fig. 6 Application of artificial intelligence in antibody mimetics
Notes: Each node represents a keyword; the size of the keyword is determined by the rate of occurrence.
Furthermore, artificial intelligence technology, such as Alphafold (Fig. 6) (Tunyasuvunakool et al., 2021 ), is applied to accurately predict the structure and conformation of an unknown protein based on regular binding and folding patterns collected from related database such as Protein Data Bank. In particular, the three-dimensional structure of an antibody can already be predicted from its gene. However, there is a big proportion of antibodies without corresponding gene available, furthermore, dynamic conformational change of antibodies when binding with antigens or receptors cannot be predicted yet (Abanades et al., 2022 ). It will greatly diminish time and cost of antibody design if protein structure could be predicted based on the structure of the antigens or receptors, while the conformational change and more limited data of the antibody-antigen complex increase the challenge for this breakthrough. However, there are still a lot of limitation of computational approach restricting its further optimization and application in antibody engineering. For example, computational analysis and AI algorithms require large amounts of data to learn and make accurate predictions. However, there is limited data available on antibody-antigen interactions, which can limit the accuracy of predictions made by these methods. Computational models used for predicting the structure and function of antibody mimetics are based on assumptions and simplifications, which may not accurately represent the complexity of real-world systems. This can lead to inaccuracies in predictions and suboptimal designs, Computational design methods often rely on the availability of templates for the design of new antibody mimetics. However, the diversity of available templates is limited, which can limit the range of possible designs.
Table 3 Antibody mimetics currently approved by the FDA or in clinical trials