Skip to content

Releases: NBDsoftware/frag_pele

FragPELE 3.1.4 Release

31 Jul 16:41
Compare
Choose a tag to compare

New minor version of FragPELE to improve integration with PELE Platform and peleffy.

FragPELE 3.1.3 Release

30 Jul 06:54
Compare
Choose a tag to compare

New minor version of FragPELE to fix a bug with external templates

FragPELE 3.1.2 Release

29 Jul 15:55
Compare
Choose a tag to compare

New minor release of FragPELE:

  • Support external templates
  • Other minor bugs corrected

FragPELE 3.1.1 Release

29 Jul 15:01
2809c07
Compare
Choose a tag to compare

Change log

  • Bux fix in Frag main

FragPELE 3.1.0 Release

29 Jul 12:56
b922769
Compare
Choose a tag to compare

Change log

  • Softcore protocol keeps charges constant in the scaffold for the first half of the growing steps.
  • New flag --keep_templates. When set it does not create and overwrite both, the scaffold and full-ligand templates.
  • PELE paths updated to 1.7.1 in MN.
  • Solvent OBC parameters automatically generated (to facilitate OpenFF simulations)
  • OFF flag: OpenFF currently working when running PELE >=1.7.1

FragPELE 3.0.0 Release

10 Jun 22:14
cbf9ba7
Compare
Choose a tag to compare

Major changes

  • Addition of growing modes. (1) Soft-like mode: charges set to 0 and included in the middle of the growing procedure. Charges are modified exponentially meanwhile other parameters grow linearly. (2) AllLinear: all parameters are grown linearly. (3) SpreadHCharge: in the first step the charge of the replaced H is spread between all atoms of the fragment. After that, the charge is replaced again for the one corresponding to each atom. The only parameters modified are VDW radii, charges and equilibrium distances, that are grown in a linear way.
  • Inclusion of constraints. PELE conf constraints can be included as arguments in the command line. F.ex: core constraints, or dihedrals constraints. To apply it on dihedrals, the user must specify which atoms of the ligand want to keep contrained. This is important to be applied when growing fragments with chiral centers.
  • New minimization protocol. In order to correct wrong poses along with the growing procedure we have strengthen the minimization to a RMS of 0.01 in the first half of the growing phase, and then it is set to 0.1 in the second half.
  • Starting from option. Now you can start simulations from an specific lambda value (from 0 to 1). The code will jump automatically to the growing step that goes immediately after the assigned lambda, and the size of the fragment in the initial step will be adapted this value.
  • Covalent Growing protocol (beta). FragPELE can now grow fragments onto amino-acids or modified residues (ligands covalently bound). A new argument and PELE conf has been included to do so. It only need the chain and the residue number to grow something onto it. Take into account that running this kind of simulations will replace the metric 'Binding Energy' for 'LocalNonBondingEnergy'.
  • Peleffy instead of PlopRotTemp. The new module PELE Force Field Yielder will be used to generate templates. It will allow us to use OpenForceField in further releases, as well including constraints in templates (fundamental for covalent ligands).

Minor changes

  • Preparation Wizard run before generating templates. As peleffy requires the connectivity to create templates, a fast preparation of ligands PDBs is done before generating templates. If this PDB file is already found, the preparation will be skipped and the code will run peleffy directly (this allows the user to correct the PDB manually if something wrong happens in the automatic way).
  • Threshold clash defaults modified. This variable is the distance (in Armstrong) to find the linking atom. If more than one atom is detected it will be considered as intramolecular clash between the fragment and the scaffold region. Now if it is set None it will use the bonding distance from the bonding distances library (atom-atom specific), but if for some reason the user wants to modify it (F.ex. sometimes wrong bonds are generated in the preparation because distances are too short) they can do it.
  • Modification of PELE control files. Small changes in some parameters of PELE conf templates.
  • Class TemplateOPLS2005 replaced by TemplateImpact. The name of the object has been changed.
  • PELE default updated. In MN now it will use the version 1.6.2_SideChainPerturbation_3. This is mandatory to grow covalent ligands.
  • Folder with aminoacid PDBs. 'frag_pele/Templates/Aminoacids' is a new directory to include PDBs of amino-acids. This is done because sometimes ProteinPreparationWizard automatically charges the atoms of the backbone and this can produce errors when comparing initial and grown templates (when growing onto amino-acids). So, if you find this problem, you can include the manually corrected PDB in this directory, and the code will try to get it to create the initial template.
  • srun as default instead of mpirun. You can choose between mpirun and srun to execute PELE with multiprocessing.
  • Code prepared to jump to OpenForceField. Everything is ready to use OpenForceField we just need to try it.
  • OFF Water template. It has been added in 'frag_pele/Templates/OFF' to copy it into DataLocal folder (temporary solution).
  • Folder handler: Protein and OFF. Folder handler can prepare now DataLocal's folders to include Protein's templates and use another force fields such as OFF.
  • Several unnecessary prints removed.
  • Small bugs fixed.
  • Test added and adapted to the new version