JabberDock provides a mechanism to dock two protein STID maps together in conjunction with the POW engine and BioBox
Note that it requires BioBOx as a means to generate the STID maps, and POW to run the optimisation.
This version of JabberDock ships with a version of BioBOx intended to allow it to work. See below for a brief introduction to installation / where to find POW.
auto_scripts Command line features are provided in the auto_scripts folder.
- Help for each one for be provided by typing the command follows by -h, e.g. dock.py -h or dock_mem.py -h
- The commands should be accessible in your PYTHONPATH
- More information on each command is provided in the accompanying manual
methods
- In here you will find all the python scripts called by the various commands in auto_scripts
- Each one has documentation written out, this can be read either directly in the file, or via the manual
- There are a few additional commands provided for users which are not used by the commands in auto_scripts...
- For example, the create_pqr_dat in data, which can convert an rtp file into a readable dat pqr file to create STID maps
JabberDock requires python 3.x (as of 09/11/20) and the following packages (and their associated packages):
- numpy
- biobox
- cython
- scipy
- scikit-image
- subprocess
JabberDock requires the following software to be installed to run:
- gromacs 5.x
- VMD 1.9.x
- POWer
See below for details on Transmembrane Protein docking requirements. Note that using versions of JD (i.e. < ver 1.1, before 09/11/20) will require Python 2.7.x
Both biobox and POWer are available with this module.
Install JabberDock with: 'python setup.py install'
After downloading from github, this JabberDock folder might be called JabberDock-master. Ensure it's called JabberDock for the imports to work: mv JabberDock-master JabberDock
Please read the full installation instructions in the manual.
Shipped with this version of JabberDock is a heavily stripped version of biobox containing the functions JabberDock requires. Please move the biobox folder into whatever directory you like (e.g. your home), just make sure it's in your pythonpath There is a readme file in the biobox folder containing instructions on requirements and installation instructions.
This is an unpubished PDB manipulation module, and therefore is unavaliable elsewhere. We are aiming to publish the full module at some point in the future.
The original version of POW can be found on the EPFL website: https://www.epfl.ch/labs/lbm/resources/
However this is an outdated version, the current version of JabberDock requires the updated version of POW (POW_v2) shipped with JabberDock.
To install, simply move the POW_v2 folder into your home (or other) directory - where JabberDock and biobox should now be located.
You will also need to install dill. To do this, type: pip install dill On your command line. For the mpi4py installation, we would recommend mpi4py 3.0.0: pip install mpi4py==3.0.0
The transmembrane protein docking of functionality is novel as of ver. 1.1. It can be used to dock two integral TM proteins by:
- Running the MD simulations in their native environment
- Restricting the search space This means the alignment and orientation of input structures is very important. The TM region centre of mass MUST be centred at the origin, with the direction of the membrane along the z axis.
TM functionality requires the additional following packages:
- Modeller (if repairing a structure)
- biopython (if repairing a structure)
- AmberTools ver 18+
Commands are very similar to the soluble protein commands, with the exception that a suffix of _mem is added (e.g. dock_mem.py) The exception is the introduction of create_mem_system.py, a new command required before gmx_run_mem.py to generate the embeded protein in a bilayer system.
For our benchmark, we used Amber14sb with SLipid forcefields. For the time being, JD membrane docking requires users to utilise their own merged forcefields (e.g. amber03_slipid default) There is quite a significant amount of defaults associated with this, mainly hacks to get the two working together. These defaults are only activated if both amber and slipid are present in your -ff flag If there is significant demand for it, I will merge and make available a viable forcefield. I am also working on an updated manual. The absence of both a merged forcefield and manual is due to pressing thesis deadlines. Note that we can't supply our Amber14sb_slipid FF as this requires an academic license.
Similar to ver 1.0, there are additional tools added to the other python modules, including chain repair functionality and reliable RMSD alignment tools which are useful to the enterprising user.
With these scripts, we are also including a folder of input structures (and associated targets) used for our benchmark.
The associated publication is currently under review at The Journal of Chemical Information & Modeling. We will link the DOI for citation once it has been through peer review.
Please see the manual in the home directory for more details on how to use JabberDock. Email the author if you have any questions.
When using JabberDock in your work, please cite the following publication: https://pubs.acs.org/doi/10.1021/acs.jctc.9b00474
This is a work in progress, and there are bound to be bugs, particularly with the variety of machines. If you have any issues, please email the author at l.s.rudden@durham.ac.uk. (Final note): I am approaching the end of my PhD (31/01/2021 thesis submission). If you like this work or any of my other publications, please get in touch with me at the above email address.