Conversation
| TAG_CHECKS = [ # Each tag should be represented here with necessary files. If no files are necessary, include an empty list. | ||
| ("protein", ["protein.pdb"]), | ||
| ("cofactor", ["cofactors.sdf"]), | ||
| ("sfe", "experimental_solvation_free_energy.json") |
There was a problem hiding this comment.
Will this work for mnsol given we can't have that file around?
There was a problem hiding this comment.
I think that means we just can not tag that dataset :(
| - mobley_7176248 | ||
|
|
||
| - am1bcc_at: | ||
| - No newline at end of file |
There was a problem hiding this comment.
Does that mean no ligands had issues?
There was a problem hiding this comment.
Ahh they just hadn't finished yet, in total 1 failed with antechamber and 1 with elf10 so not too bad!
IAlibay
left a comment
IAlibay
left a comment
There was a problem hiding this comment.
Overall this seems reasonable to me, couple of questions, and not sure about tests.
|
Depends on #98 which has modifications to the benchmark data class to capture the experimental data |
jaclark5
left a comment
jaclark5
left a comment
There was a problem hiding this comment.
Looks great! I still need to verify whether we can/can't include MNSol exp. ref data.
The way the exp data file is set up it's the ligand names, so I suppose it's implied that it's an HFE? When I update to include MNSol I can adjust logic to depend on whether a mixture_index.yaml is found containing mixture inch keys and then nothing here will break, does that sound good?
There was a problem hiding this comment.
@jaclark5 makes a good point - for consistency we should have the solvent listed in these entries.
There was a problem hiding this comment.
Yeah for MNsol we should have a solute in many solvents so how about we make the keys "solute,solvent" and have identifiers for both in the data.
There was a problem hiding this comment.
Should this file end with "data.json" as described here?
|
|
||
| The reference data was generated using the [generate_freesolv_exp_data.py](../../../data_generation/generate_freesolv_exp_data.py) script using the [conda-lock_linux-64.yml](../../../data_generation/conda-lock_linux-64.yml) environment. | ||
| Charges were generated using the [charge_freesolv.py](../../../data_generation/charge_freesolv.py) script using the [conda-lock_linux-64.yml](../../../data_generation/conda-lock_linux-64.yml) environment. | ||
| Some ligands could not be charged with all methods, the following lists the ligands that could not be charged with each method: |
There was a problem hiding this comment.
Would we want to add a ligand network for these, or not for now?
There was a problem hiding this comment.
I'll say no for now, as I don't think this will be needed yet and don't want to confuse people. We can add when needed?
# Conflicts: # openfe_benchmarks/data/benchmark_system_indexing.yml # openfe_benchmarks/tests/test_benchmark_index.py
4 participants
Fixes #55
Add FreeSolv ligands with charges and experimental data.