Skip to content

Releases: docking-org/dock3-release

DOCK3.8.5

15 Apr 20:56
@bwhall61 bwhall61
v3.8.5
4cf5b28

Choose a tag to compare

View all tags
DOCK3.8.5 Latest
Latest

DOCK 3.8.5 Release Notes
By Brendan Hall, assisted by John Irwin

DOCK Fortran (github.com/docking-org/dock3 trunk)

Fixes:

  • Fixed a bug that was doubly closing an open file. This manifested as asterisks and a 0.00 total score appearing in the OUTDOCK when using the PGF90-compiled binary on certain filesystems. “The asterisk problem”
  • Allows for user set PGF compiler path by setting PGPATH environment variable before compilation

Pydock (github.com/docking-org/pydock3 trunk)

New:

  • Added a step to pydock blastermaster which automatically generates grid visualization files to view in Chimera. NEW

Changes:

  • Changed the way matching sphere perturbation occurred in dockopt. Previously sampled perturbations in each of the x,y,z directions independently, resulting in non-uniform perturbation within a cube. Changed to use polar coordinates to more uniformly sample perturbations within a sphere.
  • Changed default matching sphere perturbation to just perturb non-xtal spheres. Added an option to additionally perturb xtal spheres.
  • Lay groundwork for individual dockopt steps to be submitted to a scheduler rather than all steps being run in sequence. Currently only implemented for the desolvation grid generation (which is one of the slowest steps)
  • Add the xtal-lig.pdb and rec.crg.pdb to the dockfiles
  • Changed the default dockopt search to just be a simple grid search
  • Updated some of the default parameters:
    • Thin spheres distance_to_ligand: 2.0 -> 4.0
    • Molecular_surface_density: 1.0 -> 10.0
    • Matching sphere perturbation max distance: 0.4A -> 0.5A
    • Extended values of the dockopt electrostatics thin sphere scan
    • Removed dockopt bump_max scan

Fixes:

  • Fixed a bug that was incorrectly creating thin spheres. The code for parsing the normal vector from dms was reading the incorrect columns.
  • Replace some hardcoded parameters with optional, configurable parameters. The default will be to use the previously hardcoded values, but they can now be changed.
  • Correctly use the molecular surface density parameters defined for electrostatic and desolvation thin spheres
  • Made makespheres1.cli.pl deterministic.
  • Improved the runtime of the function that checks for duplicate thin spheres which heavily impacted performance for higher molecular surface densities
  • Fixed an incorrect instance of a variable name being “distance_to_lig” rather than “distance_to_surface”
  • Added AMBER ff03.r1 RNA charges under pydock3/blastermaster/defaults/rna_amb.crg.oxt and pydock3/blastermaster/defaults/rna_prot.table.ambcrg.ambH

Building Pipeline (github.com/docking-org/zinc22-3d trunk)

New:

  • Added another way to submit the building pipeline with docker/apptainer containers containing all of the required software rather than unzipping the software onto /scratch every time. Also replaces many temporary files with linux piping

Fixes:

  • Added error handling in the strain calculation to ensure a vector has a non-zero length before trying to normalize it. This was causing molecules with azides to fail
  • Added the neutral smiles to the db2 file. Replaces a “fake” line with the smiles. (docker/apptainer version only for now)
  • Removed the protomer ID from the molecule name within the db2 file. Previously, protomer IDs (ex. “.0”) were included in the name, resulting in the truncation of “ZI” from the beginning of ZINC IDs.
  • Aligns the omega force field options to be consistent with those actively being used to build ZINC22

Full Changelog: https://github.com/docking-org/dock3-release/commits/v3.8.5

Assets 2
Loading