The main aim is to develop a framework that enhances antibody design through geometry-aware methods.
The framework utilizes computational geometry techniques.
It focuses on optimizing the complementarity-determining regions of antibodies.
Source code is publicly available for implementation.
GeoGAD improves precision in antibody design compared to traditional methods.
Enhanced performance in designing complementarity-determining regions is observed.
Abstract
The source code and implementation are available at https://github.com/WeiSongJian/GeoGAD, and the archival version for this manuscript is deposited at https://doi.org/10.5281/zenodo.18073443.