How to install FiltRest3D

To rank a set of decoys using FiltRest3D script you need to install required software and download the Python source.

Requirements

The software requires:

  1. Unix-like system (like Linux or Mac OS X),
  2. Python language interpreter (version 2.3 or later) and BioPython library (version 1.41 or later).
  3. Stride secondary structure assignment program;
  4. DSSP secondary structure assignment program (computes solvent accessibility),
  5. for checking consistency with Sable SS prediction you need to install Sable program;
  6. for checking knots you need the code from the paper about checking knots;
  7. for checking protein surface you need to install Situs programs qrange, pdblur, conformat and 3SOM software.

All required programs should be placed on system path, or appropriate variables need to be changed in FiltRest3D sources.

Test drive

To verify that Python and BioPython are usable with FiltRest3d, run the filtrest3d.py script without any arguments. It should produce the usage synopsis.

$ python filtrest3d.py
usage: filtrest3d.py [<options>] (-r|--restraints) <restraints filename> [<directory_names>*] [<PDB file name>*]

options:
  -h, --help            show this help message and exit
  -r RESTRAINTS_FILE, --restraints=RESTRAINTS_FILE
                        File with restraints description
  --limit=N             process at most N structures
  -d DIRFILE, --dirfile=DIRFILE
                        Path to file with list of PDB files, one path per
                        line.
  --silent=SILENTFILE   Path to Rosetta silent file ('.out' or '.out.gz') with
                        structures (contains just C-alpha + SS assignment
  --traf=TRAFFILE       Path to CABS TRAjectory file ('.tra' or '.traf')
  --seq=SEQFILE         Path to CABS SEQuence file complementing TRAjectory
                        information
  --fasta=FASTAFILE     Path to FastA sequence file complementing TRAjectory
                        information
  -l, --linear          Use linear punishment function for distance
                        restraints.
  -s, --sqrt            Use square punishment function for distance
                        restraints.
  --knotted             Assume protein has a knot.
  -k, --not-knotted     Assume protein has no knots.
  --check-solvation-prediction
                        Assume protein has no knots.
  --verbose             Print debug information.
  --debug-memory-use    print number of live objects after processing each
                        file
  -p PYMOL_SCRIPT, --pymol-script=PYMOL_SCRIPT
                        Output PyMol script for restraints visualization
  -i, --incremental     Print results as they are generated
  --model-pickle=MODEL_PICKLE
                        filename to put the pickle of PDB
  --run-pymol           run PyMol on first model and generated PyMol script
  --progress-bar        print dot for each parsed model

To check that Stride and DSSP programs are properly installed, run the FiltRest3D on two sample files attached with the distribution: cleaned version of native structure 1GEF and restraints derived from the paper on the same protein.

$ ./filtrest3d.py --restraints example/1gef/1gefA_restraints.filtrest example/1gef/1gefA.pdb
1 structures
filename : sum : D33_near_K48 : beta1_near_beta2 : E9_near_D33 : E9_near_K48 : D33_near_E46 : beta3 : E9_near_E46 : beta1 : beta2_near_beta3 : E46_near_K48 : beta2
example/1gef/1gefA.pdb : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0 : 0.0
Copyright '2006 by Michal J. Gajda