search

openforcefield / openff-benchmark

Comparison benchmarks between public force fields and Open Force Field Initiative force fields
MIT License
11 stars 2 forks source link

Hydrogen indexing issue when exporting from QM jobs #42

Closed j-wags closed 3 years ago

j-wags commented 3 years ago

Sometimes I get results like this when exporting from QM compute:

Screen Shot 2021-01-25 at 10 13 08 AM

The geometry is reasonable but the connectivity table (at least as pymol sees it) might be mangled.

This is particularly confusing because both initial and final MM energies (during a minimization that STARTS from this structure) are reasonable: 

>  <initial_energy>  (1) 
-13.078687136788444 kilocalorie / mole

>  <final_energy>  (1) 
-23.499628545727884 kilocalorie / mole

If the indexing really were mangled I'd expect the initial energy to be very high.

jthorton commented 3 years ago

I think the cause of this is the use of Molecule.from_smiles here instead of Molecule.from_mapped_smiles this code path is only followed however if users do a direct optimization using qcengine and skip the qcarchive server which should only effect compute procedures C and E.