Traditional approaches for antibody structure prediction like X-ray crystallography and NMR are time-consuming and does not support high-throughput pipeline. In-silico based approaches can solve these problems by speeding up the process and High-throughput screening.
I researched and adopted existing approaches for structure prediction from amino-acid sequences. I also learned about Homology Modelling and developed a pipeline for the same. A new approach that I tried was by combining both Homology Modelling using RosettaCommons and deep learning approach to get the antibody structure.
The implementation code is propriety property of Aganitha Cognitive Solutions.