Hi Rocco, Steven, et al,
Posting this here since it appears that the Rosetta bugs reporting tool is currently down.
Several months ago I found some bugs in the HPatchIG code that crept in from RotamerSet* and sasa scoring refactoring, which the HIG unit/integration tests unfortunately did not catch. I emailed the bug and a patch file for the fix to Andrew, but I don't think he was able to get the fix in before leaving, as they are still present in the weekly release (2022.40, release 331) that was recently posted to the downloads page. To get back to a working state, three lines need to be changed in src/core/pack/interaction_graph/HPatchInteractionGraph.hh. The diff (to the latest weekly release) is as follows:
835c835
< chemical::ResidueType const & ii_restype = rotamers_vector_[ rotamers->get_residue_group_begin( ii ) ]->type();
---
> chemical::ResidueType const & ii_restype = rotamers_vector_[ rotamers->get_residue_type_begin( ii ) ]->type();
3424c3424
< total_sasa = core::scoring::calc_per_atom_sasa( pose_copy, atom_sasa, rsd_sasa, probe_radius, false /* no big polar H */, atom_subset, true /* use_naccess_sasa_radii */, true /* expand polar radii */ );
---
> total_sasa = core::scoring::calc_per_atom_sasa( pose_copy, atom_sasa, rsd_sasa, probe_radius, false /* no big polar H */, atom_subset, true /* use_naccess_sasa_radii */ );
3939,3940c3939
< //conformation::ResidueCOP rsd( ii == node_considering_alt_state_ ? get_hpatch_node( ii )->curr_state_rotamer() : get_hpatch_node( ii )->alt_state_rotamer() ); // old bug, only rarely encountered
< conformation::ResidueCOP rsd( ii == node_considering_alt_state_ ? get_hpatch_node( ii )->alt_state_rotamer() : get_hpatch_node( ii )->curr_state_rotamer() );
---
> conformation::ResidueCOP rsd( ii == node_considering_alt_state_ ? get_hpatch_node( ii )->curr_state_rotamer() : get_hpatch_node( ii )->alt_state_rotamer() );
Could one of you please submit a PR for these fixes? Thanks in advance.