DUMM6

From DISI
Jump to navigation Jump to search

Attendees & Time

Sarah, Michael M, Ryan, Magdalena, Peter, Kong, Jens, Dahlia, John

Date: April 2nd, 2010. 1:05 pm.

Slides

1. Philosophy

  • deleted unused things
  • cleaned up code
  • streamlined call graph
  • single mode gone (use version 3.5.54.70)
  • contact scoring gone
  • precalculate various grid terms

2. DOCK 3.5.54.72 improvements

  • chem.bmp bumping instead of distmap contact bumping
  • new hierarchy code (should eliminate some broken)
  • no more out of bounds molecules
  • no more clashing molecules
  • deprecated unused keywords (more later)
  • vdw repulsive term capped at a million (no more overflow bugs with huge negative energies)
  • 64bit/32bit builds (use dock.csh instead of calling directly)
  • speedups (3-5x)

3. Deprecated keywords

  • write_coordinates
  • standard_pdb
  • focus_ratio
  • normalize_save
  • restart_interval
  • mirror_ligands
  • scoring_option
  • mode
  • ligand_type
  • restart
  • output_hydrogens
  • recombine_fragments
  • interpolate
  • ligand_sphere_file

4. New keywords

  • keyword default suggested
  • remove_positive_solvation no YES
  • check_clashes yes YES

5. DISI

  • it is great


All of this is live since Mar 18 2010 in dock72!

Hardware updates

Software updates

Requests/Improvements

Sarah raised the issue of the flexible protein code of Niu Huang:

  • Is it compatible with the current improvements? <math>\rightarrow</math> probably not.
  • Is anybody using it?
  • Would it be worth it to update and include it?
  • Can it be used in an automatic fashion, e.g. for DOCK Blaster?
  • How can we predict water positions and usefulness (maybe initial water placement with PLOP)?


<math>\Rightarrow</math> who will be doing all of this?

Database updates

Varia

Errata

Action Items

→ back to DUMM main page