-
Notifications
You must be signed in to change notification settings - Fork 23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Chemically equivalent atoms with delocalized formal charge are assigned different torsions #56
Comments
To add to the list, similar issue with Nitro groups,
|
Whatever's going on, I can verify it's not due to recent changes in the toolkit. The reproduction behaves the same back to 0.9.0 and modifying it to the |
@mattwthompson Yeah, it's not toolkit related, it's from the smarts definition of applied parameters distinguishing single and double bonds. |
Copying from slack discussion: introducing new parameters t18a, t83a, t87a, which would explicitly match these cases can be a potential workaround.
|
Following up on some suggestions from Chris Bayly, it looks like phosphate esters and sulfate esters are assigned correctly because there is only one torsion SMIRKS that matches them. The nitrogen torsions in the amidinium cation are assigned incorrectly.
produces
|
Describe the bug
Chemically equivalent atoms in which a nonzero charge is delocalized (represented by resonance structures with the formal charge on different atoms) are assigned different torsions by OpenFF 2.0.0. Examples include carboxylate and guanidinium groups.
To Reproduce
MWE for the amino acids aspartate (carboxylate) and arginine (guanidinium).
Output
The code above produces
For aspartate, the C-C-C=O torsion gets t18, but C-C-C-[O-] gets t17.
For symmetric arginine, where the formal charge is on the epsilon nitrogen, both C-[N+]=C-N torsions get t87.
For asymmetric arginine, where the formal charge is on one of the zeta nitrogens, the C-N-C=[N+] torsion gets t79, but C-N-C-N gets t75.
Computing environment (please complete the following information):
OS is Ubuntu 20.04.5.
openff-toolkit
is 0.10.6.openff-forcefields
is 2.0.0.Output of
conda list
isThe text was updated successfully, but these errors were encountered: