ParamChem was a NSF sponsored initiative to develop an integrated cyber environment to address the simulation needs of molecular sciences. While it ran out of funding before meeting its ambitious end-goal of releasing a fully-integrated GUI-driven infrastructure for automatic parameter optimization, the technologies developed as part of the project represent substantial advancements in the field; please visit our technology page to access the CGenFF program, the lsfitpar parameter fitting software etc. All text below this point pertains to the original project that is now defunct, and only serves historical purposes.
The proposed infrastructure will provide reference data organizers and generators as well as workflows for automatic parameterization of Molecular Mechanics (MM) Force Fields as well as Semi-Empirical (SE) methods.
The major goals of the project are
- A comprehensive utility for the optimization and testing of parameters in Force Fields and Semi-Empirical models
- Utilities for systematic expansion of available parameter sets to novel chemical systems
- A searchable database of target data for parameterization as well as parameter sets
The utility mentioned in the first goal will be primarily useful to experts, allowing them to develop novel models of higher accuracy in shorter time periods. These models can then be made available to the computational chemistry community at large via the database mentioned in the second goal. This will make it easier for computational chemists to find an appropriate model for the system they are studying, and, if necessary, to extend the model to novel functional groups using the utilities mentioned in the third goal.
Many existing popular Molecular Mechanics and Semi-Empirical models will be integrated. From this, a wide range of parameters encompassing biological, organic and inorganic species will be accessible for direct use or further optimization. The infrastructure will be extensible in terms of data sources and energy functions allowing for its applicability in any parametrization scheme.