- Data Assimilation with Single-Molecule Measurements
- Dimensionality Reduction Using UMAP
- ML/MM for Enzymatic Reactions
- Backmapping Method using Diffusion Models
Research
AI and Machine Learning
AI and machine learning methods, as exemplified by AlphaFold, possess high predictive capabilities. However, the predictive performance of machine learning depends heavily on the quality and quantity of the data used for training. By combining machine learning with multiscale molecular dynamics (MD) or by employing techniques that assimilate experimental data (data assimilation), it is possible to achieve even higher-precision predictions on biomolecular dynamics related to their functions.
Going forward, we will revolutionize multiscale MD by utilizing machine learning algorithms (such as CGBack and ML/MM). We will also launch new research initiatives, such as storing MD simulation results in a database and using that data in machine learning.
Methods Developed to Date
