DMFF v1.0.0 Released!
New features:
- Implemented QEQ electrostatic potential under constant charge and constant potential modes.
- Refactored the API layer, enhancing frontend robustness.
- Support calculating electrostatic potential in non-periodic systems for point charge, BCC charge, QEQ model, and multipole model.
- Added dmff.api.DMFFTopology class for storing topological information.
- Added dmff.api.ParamSet class, supporting Pytree data structures for improved management of force field parameters.
- Improved computation efficiency of PME kernel on GPU.
- Implemented backend/save_dmff2tf.py, which stores DMFF potential as a TensorFlow module for use in C/C++ API calls.
- Developed an OpenMM plugin, enabling the use of DMFF potential within OpenMM.