Releases: openforcefield/openff-forcefields
OpenFF 3.0.0 alpha 0 (no water model)
This release adds openff_no_water-3.0.0-alpha0.offxml and openff_no_water_unconstrained-3.0.0-alpha0.offxml. This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, and the monoatomic ions Li+, Na+, K+, Rb+, F-, Cl-, Br-, and I-.
We’re making an early alpha version of a SMIRNOFF protein force field available. While this FF has passed our initial rounds of benchmarks, there are more rounds to evaluate, and our previous experience leads us to expect that this will not be the FF that we release as Rosemary (openff-3.0.0). However we wanted to give users and developers an early look to begin their planning. We anticipate that the full openff-3.0.0 release will have the same structure as this alpha protein FF, in terms of using AshGC/NAGL charges (not library charges) and the approximate number and composition of SMIRKS parameters.
This force field is a research tool and does not contain a water model, and probably has other pointy edges. It’s likely you’ll get the best results using OPC3 as the water model. If you do not include a water model, the FF will still work, but your waters will be assigned hydroxyl parameters, will lack angle constraints, and will be very unphysical.
Limited training data and code is available for this alpha force field at https://zenodo.org/records/17401377 . This is the DOI we'd prefer researchers use to cite this FF.
Sage 2.3.0 Release Candidate 2
This release adds openff-2.3.0-rc2.offxml and openff_unconstrained-2.3.0-rc2.offxml. Sage 2.3.0 is the first OpenFF force field to use the AshGC neural network charge model to assign charges. Both vdW parameter values and valence parameter values have been re-fit. Compared to 2.2.1, it splits torsions so that a single torsional parameter only covers a central bond with a single multiplicity. It also adds some bond and angle parameters to fix targeted chemistries.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, atomic Xe, and the monoatomic ions Li+, Na+, K+, Rb+, Cs+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this force field also contains parameters for TIP3P, including bond length constraints.
Sage 2.3.0 Release Candidate 1
This release adds openff-2.3.0-rc1.offxml and openff_unconstrained-2.3.0-rc1.offxml. Sage 2.3.0-rc1 is the first OpenFF force field to use the AshGC neural network charge model to assign charges. Both vdW parameter values and valence parameter values have been re-fit. Compared to 2.2.1, it splits torsions so that a single torsional parameter only covers a central bond with a single multiplicity. It also adds some bond and angle parameters to fix targeted chemistries.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, atomic Xe, and the monoatomic ions Li+, Na+, K+, Rb+, Cs+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this force field also contains parameters for TIP3P, including bond length constraints.
Sage 2.2.1
This release adds openff-2.2.1.offxml and openff_unconstrained-2.2.1.offxml, Sage 2.2.1, which is identical to the Sage 2.2.1 release candidate. Compared to Sage 2.2.0, it fixes some linear angles to stay at 180 degrees. Code, environments, and data used to fit this force field can be found here.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, atomic Xe, and the monoatomic ions Li+, Na+, K+, Rb+, Cs+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this force field also contains parameters for TIP3P, including bond length constraints.
Sage 2.2.1 Release Candidate 1
The new force field files in this release are from 2.2.1-rc1 tag of the Sage-2.2.1 repo.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, and the monoatomic ions Li+, Na+, K+, Rb+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this contains parameters for TIP3P, including bond length constraints.
Sage 2.2.0
This release adds openff-2.2.0.offxml and openff_unconstrained-2.2.0.offxml, Sage 2.2.0, which is identical to the Sage 2.2.0 release candidate. Compared to Sage 2.1.0, it modifies some small ring internal angles and remedies issues with sulfamide geometries. Code, environments, and data used to fit this force field can be found here. Attached to that release is the release tarball, Source code, which is a static snapshot of all the code and data used to fit and benchmark the force field. The fitting process should be reproducible from this snapshot. The openff-2.2.0.offxml force field released here is identical to the file in that repo.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, atomic Xe, and the monoatomic ions Li+, Na+, K+, Rb+, Cs+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this force field also contains parameters for TIP3P, including bond length constraints.
Sage 2.2.0 Release Candidate 1
This release adds openff-2.2.0-rc1.offxml and openff_unconstrained-2.2.0-rc1.offxml, Sage 2.2.0 Release Candidate 1, which modifies some small ring internal angles and remedies issues with sulfamide geometries. Code, environments, and data used to fit this force field can be found here.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, atomic Xe, and the monoatomic ions Li+, Na+, K+, Rb+, Cs+, F-, Cl-, Br-, and I-. The parameters have been co-optimized to TIP3P water, therefore this force field also contains parameters for TIP3P, including bond length constraints.
Sage 2.1.1
This release adds openff-2.1.1.offxml and openff_unconstrained-2.1.1.offxml, Sage 2.1.1, which has identical parameters to Sage 2.1.0 (originally released in version 2023.05.1 of the openff-forcefields package) but adds Xe van der Waals parameters from Tang, K.T., Toennies, J.P. New combining rules for well parameters and shapes of the van der Waals potential of mixed rare gas systems. Z Phys D - Atoms, Molecules and Clusters 1, 91–101 (1986).
The new OFFXML files in this release were created by running scripts/add_xe_vdw.py, available in the source tarball.
This force field is applicable to druglike molecules consisting of the elements C, H, O, N, P, S, F, Cl, Br, and I, and the monoatomic species Xe, Li+, Na+, K+, Rb+, F-, Cl-, Br-, and I-. The non-monoatomic parameters have been co-optimized to TIP3P water, therefore this contains parameters for TIP3P, including bond length constraints.
Adding more water models
This release adds the TIP4P-EW, TIP5P, and SPC/E water models. There are also some minor code changes to ensure compatibility with Python 3.12
What's Changed
- Bump codecov/codecov-action from 3 to 4 by @dependabot in #81
- Bump actions/checkout from 3 to 4 by @dependabot in #80
- Revert "Bump codecov/codecov-action from 3 to 4" by @j-wags in #82
- Remove use of
pkg_resourcesby @mattwthompson in #78 - Test virtual site weights in water models by @mattwthompson in #87
- Add TIP4P-EW by @mattwthompson in #76
- Add SPC/E by @mattwthompson in #75
- Add TIP5P by @mattwthompson in #85
- Update versioneer setup by @mattwthompson in #91
- Remove more uses of
pkg_resourcesby @mattwthompson in #90 - Update virtual site-related tests by @mattwthompson in #92
- Update
setup.pyby @mattwthompson in #88 - Uniformly lint by @mattwthompson in #89
- Do not yet enforce EOF consistency by @mattwthompson in #93
Full Changelog: 2023.08.0...2023.11.0
Updating water models with vdW 0.4
This release updates water models to use version 0.4 of the vdW section of the SMIRNOFF specification.