Conclusion

With this new approach we were able to capture approximately 60% of the varience using strucutral alone.

As documented in the paper, we attempted to use PCA with VCD calculations and 3rd party embeddings as features to the model to improve performance. Niehter approach was better at capturing varience than the structural model.

We believe the remaining varience can be explained by a better understanding of the ligand-receptor binding process.