ProtScape: A molecular structure and energy-aware representation for protein conformation generation
ProtSCAPE is a deep learning model designed to understand protein dynamics from molecular dynamics simulations. It uses a novel architecture combining geometric scattering transforms and transformer attention mechanisms to create latent representations of protein structures. This allows the model to capture how proteins change conformation over time, enabling predictions of flexibility and structural transitions.