YabXnization platform: A monoclonal antibody heterologization server based on rational design and artificial...

Area of Science:

• Immunotherapy
• Protein Engineering
• Computational Biology

Background:

• Antibody therapeutics are vital in immunotherapy.
• Antibody heterologization (e.g., humanization) is essential before in vivo application.
• Existing methods can be complex and time-consuming.

Purpose of the Study:

• To introduce YabXnization, a novel platform for antibody heterologization.
• To enable efficient antibody engineering using rational design and AI.
• To improve the success rate and preserve binding affinity of engineered antibodies.

Main Methods:

• YabXnization employs two heterologization approaches: CDR-grafting with backmutation and AI-assisted computational design.
• AI-assisted design utilizes evolutionary computation, a DeepForest-based humanness model, and distance metrics.
• The platform was validated using 18 antibodies for humanization, caninization, and felinization.

Main Results:

• The YabXnization platform achieved a 90% success rate in antibody heterologization.
• Engineered antibodies showed binding affinity loss within one order of magnitude compared to chimeric controls.
• Some heterologized antibodies exhibited enhanced binding affinity.

Conclusions:

• YabXnization offers an efficient and effective solution for antibody heterologization.
• The platform integrates rational design and AI for robust antibody engineering.
• This technology holds significant promise for advancing immunotherapy development.