CG2AA:
Backmapping Protein Coarse-Grained Structures
L. E. Lombardi (IC/DM) - M. A. Martí (IQUIBICEN/DQB) - L. Capece (INQUIMAE/DQIAQF)
Coarse-grained models allow long-scale simulations with a much lower computational cost than that of all-atom simulations.
However, the absence of atomistic detail impedes the analysis of specific atomic interactions that are determinant in most interesting processes in biomolecules. In order to study these phenomena, it is necessary to reconstruct the atomistic structure from the coarse-grained representation.
This structure can be analyzed by itself or be used as an onset for atomistic molecular dynamics simulations.
Here we present an algorithm that accurately reconstructs the atomistic structure from coarse-grained models from the force field developed by
Hills, Lu and Voth , using a simple geometrical algorithm.
Download
The software, wich consists in a python script, is freely available here:
cg2aa.py. It requires the NumPy library. It is likely you have already everything installed for cg2aa to run.
If not, in Ubuntu or Debian, for example
apt-get install python2.7 python-numpy
should do the trick.
For more information and installation guide for the requiered software, please visit
Python 2.7 and
NumPy.
Tutorial
In order to run the program, you need a PDB file with your protein structure in the Coarse Grained representation.
The name of this structure should end in 'cg.pdb'. Now, to convert a file named 'NAMEcg.pdb', type:
/path_to_script/cg2aa.py path_to_file/NAME
For example, if the current directory holds not only the script but the file to process, just type:
./cg2aa.py NAME
The output will be an all-atom pdb file named 'NAMEbm.pdb'.
Please, note that you can provide a reference all-atom pdb file to recall histidine information by placing a file named NAME.pdb in the same folder. In that case, histidines are going to be labeled as HIE, HID or HIP as in the original all atom pdb.
If the reference pdb is not present, all histidines will be named HIS.
The reference structure can be used to get a root mean squared deviation report (of the backmapped structure with respect to the original one) by adding the -rmsd flag:
./cg2aa.py -rmsd NAME
A modest help is available via the -h or --help flags:
./cg2aa.py --help
Example: Myoglobin (PDB entry 1VXD)
Start with the coarse grain file
Mbcg.pdb.
From the comand line, type
path_to_script/cg2aa.py path_to_file/Mb
For instance, if both the script and the file are in the current directory, type:
./cg2aa.py Mb
You sould get a file named
Mbbm.pdb.
If you want, you can provide a reference
Mb.pdb for histidine labeling.
In that case, the output should look like
this.
Here, you can visualize the results of this process: the original all atom pdb in green and in cyan the output of backmapping the coarse grain structure.
Visualization of the CG model
We provide a script:
tcl-gen.py
that generates a structure-specific tcl code to run inside
VMD
in order to present a custom visualization. Its usage is simple, from a command line just type:
./tcl-gen.py CGFILE.pdb
An it will produce a tcl script named CGFILE.tcl. For the example above, for instance, the command
./tcl-gen.py Mbcg.pdb