PELE Platform 1.6¶
The new release of the PELE Platform contains many new features and stability changes. The most important ones are:
Fragment libraries: provide us with a PDB or SD file containing your fragments library and we will enumerate all possible ligands, grow them inside your binding site and rank them from best to worst, always taking receptor flexibility into account.
Conformation perturbation: if you provide PELE with a conformation library, it will perform an additional perturbation step focused on changing the conformation. Especially useful if you want to focus on bioactive conformers only or to apply specific sampling to tricky compounds like macrocycles or heterocyclic compounds.
Brand new analysis: we give users more flexibility and tools to simplify the analysis of PELE simulations. You can now choose your preferred clustering method, its parameters and criteria to extract most promising structures. We also improved the plots and added information about the water clusters resulting from aquaPELE simulations.
Covalent docking: we have developed a covalent docking protocol to let you explore all those irreversibly binding ligands. All you have to do is provide the platform with a ligand covalently attached to the required protein residue and it will sample local conformations to optimize the interactions for you.
New built-in Plotter: a new plotter to quickly represent the results of your simulation using the command line. You can choose from a number of customizable scatter, density and interactive plots.
Interaction restrictions: with the new release, you will be able to restrict the interactions between the ligand and the protein by specifying a custom H-bond distance or angle range between selected atoms.
Open Force Field integration: from now on you can parametrize your ligands and cofactors using Open Force Field as well as the good old OPLS2005.
Singularity: from now on, we support launching PELE from Singularity containers to save some of the trouble with dependencies and compatibility.
Other changes are:
Kernel density estimator plots
Constraints levels for alpha carbons
Changes to Monte Carlo parameters
Allosteric package renamed to site_finder
Minor changes to folder structure
Interstep logger integration
New optional flags to control top clusters and their representative structures selection
Recover restart flag to allow the users to manually curate control files and use them in a new run
Support for the new equilibration mode flag of Adaptive
Fix problems with the box of the Out –> In package
Add user warnings to facilitate the system preparation
Add water sites analysis
New environment variables PELE_EXEC, PELE_DATA and PELE_DOCUMENTS
Fix for the site_finder randomization
Support for a new PELE flag called minimum steps
Improvements in the docs
New checker for flag incompatibilities
New checker for input PDB files
Multi representative structures option
Fix for atom_dist bug in biased simulations
High-throughput fragment screening
Improved out-in exploration
Support for non-standard residues
Automatic metal constraints
Outliers removed from plots
Site finder (pocket exploration) package
GPCR orthosteric package
External metal constraints
Add n random water to your simulation by setting n_waters flag
More robust error handling
Remove support python 3.6 and update features python 3.7
Full refactor of code
Improvement of frag_pele
Coverage up to 94%
Include further testing of alignment and rdkit symmetry problem
Include more flags for FragPele
Improve exceptions with custom errors
Fix rdkit substructure search symmetry problem by alignment
FragPELE better tested
Coverage Platform up to 90%
Pyyaml checker for unexisting keywords in input
Improve substructure search on symmetric cases
FragPELE supported (Beta-version)
PPI simulation supported. Global exploration + induced fit (Beta-version)
Make Platform work through SCHRODINGER and PELE environment variables
Get rid of PyMol as external dependency
Use can define several inputs with asterics. i.e. “complex*.pdb”
Fix bug on dimer constraints only detecting one chain
Fix other minor bugs
Better coverage (77%)
Make mae flag convert clusters as well as top poses to mae
Let user choose number of clusters through analysis_nclust flag
Allow user to specify the columns of the report via be_column, te_column and limit_column.
Include only analysis flag
Automatically score the simulation by making the average of the 25% best energy structures.
Reorder top energy structures
Support conda deployment for python 3.8
Fixed bug in xtc analysis
Renew environment on SCHRODINGER subprocess
Set constraints by smiles
Include a default posprocessing module with plots, top poses and clusters
Separate between AdaptivePELE induced fit (induced_fit_fast) and PELE indeced fit (induced_fit_exhaustive)
Include skip_ligand_prep option to jump PlopRotTemp missing residue
Give option ot the user to specify the atom_dist by chain:resname:atomname (A:125:CA)
Give option mae to transform the best structures to mae files with the metrics as properties
Fix minor bugs
Automatic PCA mode
Fix minor bug on global exploration
Set PPP as external dependence
Fix global exploration bug when joining ligand & receptor
Add rescoring feature to local a single minimum
Add induce_fit mode and exploration mode within water_lig parameters to explore hydration sites without moving the ligand or while making the entrance of the ligand.
Some minor fixes
Add waterPELE and set defaults as we did on WaterMC paper
Include executable path, data and documents overwriting all constants.py
Insert AdaptivePELE as external dependency
Fix minor bugs
Automatic Platform to automatically launch PELE&adaptivePELE. It creates the forcefield parameters, the control files, the PELE input.pdb and finally launch the simulation.
Flexibility to include MSM and Frag PELE
Flexibility to include analysis scripts
Flexibility to include PELE modes