Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
fix mismatch between intra and unbound energy
This bug did not affect poses but it affected reported energies. In Vina (not in AD4), the UNBOUND is INTRA of the first pose. - macrocycle glue potential was added to UNBOUND but not to INTRA in the output pdbqt (during the search it added to the cost function as it should be). This would cause discrepancies of about 10 kcal/mol per glue pair. - poses were not resorted after refining with non_cache, allowing the UNBOUND energy to be the INTRA from a pose other than the first. This affected many poses by a few tenths of a kcal/mol, and some poses by a few kcal/mol. - After refinement with non_cache, it was possible to have movable atoms just outside the grid box because of a positive tolerance margin set in non_cache.h. The out-of-box penalty used during refinement is 100/angstrom in the first refinement iteration, but the unbound (intramolecular) energy is calculated with 1,000,000/angstrom. Thus, in rare cases, flex sidechain atoms would be refined to be just outside the grid box with a tiny out-of-box penalty, then sorted using the tiny penalty, and finally rescored with an UNBOUND energy using the large penalty, leading to UNBOUND that can reach 100 kcal/mol.
- Loading branch information
c1217c1
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@diogomart Very happy you found this and commit is good 👍