Add results from the openmm840 qa#117
Conversation
|
TODO:
|
jaclark5
left a comment
jaclark5
left a comment
There was a problem hiding this comment.
Here are some suggested changes to be consistent between the submission.yaml and the prepare_metadata_submission.py
|
Update _example_plan_rbfe.py to make sure that the "system_name" and "system_group" are saved in the mapping annotations. |
|
I think I updated everything according to the meeting today.
|
This should be resolved |
| openmm_version: TODO | ||
| openff_toolkit_version: TODO | ||
| mapper: "KartografAtomMapper 1.2.0 (LSA)" | ||
| forcefield: ["ff14SB", "phosaa10", "tip3p_HFE_multivalent", "tip3p_standard"] |
There was a problem hiding this comment.
This is listed here and in the protocol settings is this the searchable tag? If so we probably want the small molecule forcefield here as well.
There was a problem hiding this comment.
Or is the tags list intended to be the searchable keys?
There was a problem hiding this comment.
The tags list are intended to be searchable keys generally, but we could also remove the ff from there and expect it to be searchable from this field only, what do you think?
Gufe seems to separated the small molecule force field settings from forcefield which is the protein and solvent force field in OpenMMSystemGeneratorFFSettings. I did neglect to add the small molecule force field here. I'll resolve that
| This submission describes the charge_annihilation_set, jacs_set RBFE benchmark | ||
| (charge_annihilation_set: egfr, irak4_s2, irak4_s3; jacs_set: p38, tyk2) prepared with | ||
| tip3p_HFE_multivalent/ff14SB/phosaa10/tip3p_standard and am1bcc_at. The submission contains 160 | ||
| edges, 45 unique ligands, and 1 unique proteins. Note this means the charge annihilation sets are |
There was a problem hiding this comment.
1 unique proteins
This is wrong
There was a problem hiding this comment.
Good catch, it's because we don't set the component name or molprops, only the atoms and bonds are distinguishing features. @jthorton do you have a standard way of distinguishing proteins? If not, I'll figure something out.
| summary: | | ||
| This submission describes the charge_annihilation_set, jacs_set RBFE benchmark | ||
| (charge_annihilation_set: egfr, irak4_s2, irak4_s3; jacs_set: p38, tyk2) prepared with | ||
| tip3p_HFE_multivalent/ff14SB/phosaa10/tip3p_standard and am1bcc_at. The submission contains 160 |
There was a problem hiding this comment.
Force fields are listed again but not the small molecule force field or charges.
| if protocol_settings.partial_charges: | ||
| _add_value_with_keys( | ||
| self.partial_charges, | ||
| protocol_settings.partial_charges, | ||
| keys, | ||
| ) |
There was a problem hiding this comment.
This pulls charges from the protocol; if there are charges on the ligand, they take priority. In this case, it reports at_am1bcc (the default), but the submission used nagl. We should make this blank so the user has to fill it in.
3 participants
A draft of the results object from a local qa run of openmm-8.4.0.