Visual Analytics for Deep Embeddings of Large Scale Molecular Dynamics Simulations

摘要: Molecular Dynamics (MD) simulation have been emerging as an excellent candidate for understanding complex atomic and molecular scale mechanism of bio-molecules that control essential bio-physical phenomenon in a living organism. But this MD technique produces large-size long-timescale data are inherently high-dimensional occupies many terabytes data. Processing immense amount meaningful way is becoming increasingly difficult. Therefore, specific dimensionality reduction algorithm using deep learning has employed here to embed the lower-dimension latent space still preserves inherent characteristics i.e. retains biologically information. Subsequently, results embedding models visualized model evaluation analysis extracted underlying features. However, most existing visualizations embeddings limitations evaluating We propose interactive visual analytics system simulations not only evaluate explain but also analyze various simulations. Our enables exploration discovery semantic supports by quantitatively described features (even without labels).