TransAtlas:

Atlas of Protein Conformational Transitions

TransAtlas Targeted MD Tutorial

TransAtlas database contains more than 8 million different atomistic structures reconstructed from the intermediate frames obtained for the set of conformational transitions computed (see Atomistic Analysis section). These intermediate atomistic structures can be used as input for further biased simulations. This biased simulations can provide important information on the conformational transition path and the protein flexibility and energy required to acquire the final conformation. One of the most well-known software to compute biased MD simulations is Plumed. Plumed package is an open source library for free energy calculations in molecular systems which works together with some of the most popular molecular dynamics engines such as Amber, Gromacs and NAMD. In this tutorial, the necessary steps to go from an atomistic intermediate structure (origin) provided by TransAtlas database to another one (target), using the method called Targeted Molecular Dynamics are presented.

The first step in this tutorial is to choose the intermediate atomistic structures (origin and target) we are interested in. For that, we need to go to the Atomistic Analysis tab in the TransAtlas results.

To reach this Atomistic Analysis tab, we need to first choose a particular conformational transition from TransAtlas database. For that, we can use any of the search possibilities offered by the web server (see Search help sections). In order to use a quick and easy example to ilustrate this step, we can use the Quick Search box placed in the top-right part of the web server:

If we write the PDB code 1ba2 (a mutant of a d-ribose-binding protein, well-known to be an open state of a conformational transition) in the Quick Search box and type the return key, TransAtlas interface will send us to the results tab of the Search section:

Once there, we are going to choose the first result, the one corresponding to the conformational transition cluster representative 1ba2_A to 1urp_A:

From the different information found in the results box, the small grey box with icons inside is the one that links the conformational transition to the set of TransAtlas analysis. In particular, clicking on the magnifier icon will send us straight to the Atomistic Analysis:

Now, we can already choose a the atomistic intermediate structures to be used as origin and target from the set of 5 frames (in this particular case, the number of intermediate structures varies with every conformational transition). To select a particular originstructure, the origin white square on the slider (that is placed on the frame number 1 by default) can be moved to the right. When the slider is moved, the corresponding structure box placed just below is automatically modified with the new frame number, the difference in RMSd from the origin and target structures, the structure to be used in the visualization with JSMol and to be downloaded, and the atomistic MD setup with MDWeb. Similarly, the target structure can be selected using the target white square on the slider (placed on the last frame by default).

Once the origin and target desired intermediate structures are selected, the necessary files to start with the biased TMD simulation can be downloaded clicking at the Generate button of the Targeted MD box (bottom of the page, Path Finding Techniques section). Before clicking on the button, please check the frame numbers for the origin and the target structures shown in the box. The TransAtlas generated tgz package contains 4 different files:

  • Origin Frame: PDB file corresponding to the origin frame selected in the atomistic intermediate structures slider.
  • Target Frame: PDB file corresponding to the target frame selected in the atomistic intermediate structures slider.
  • Plumed input configuration file: DAT file containing the minimum number of commands for plumed software to compute the biased TMD simulation.
  • Auxiliar Perl script renumCaIndex.pl: PERL script to automatically reassign the same atom indexes in both starting and origin structure. Note that this is extremely important for the simulation to correctly run, and it is usually needed because the MD setup process required for the origin structure typically modifies the atom indexes, mainly because new atoms (hydrogen atoms) are included.

The second step in this tutorial is the setup of the atomistic origin structure to be used in the biased TMD simulation. For that, the MDWeb tool that is linked to TransAtlas (see Atomistic MD Tutorial section) can be used. Otherwise, the users can just apply their own set of MD setup scripting tools adapted for a desired MD package. The third step involves the recently built structure prepared to run the MD simulation (in PDB format) and the target structure found in the tgz package. First of all, the origin and target structures must be structurally aligned before starting the biased simulation. That is extremely important because the 3D coordinates are the ones used to drive the structure from one conformation to the other. Again, the MD setup process can change the initial coordinates of the origin structure, and the target one must be re-aligned accordingly. This can be easily done with different bio-structural programs, such as PyMol, Kabsch or VMD. After the structural alignment, and as explained earlier in this tutorial, we must ensure that the two structures have the same atom indexes so that plumed can correctly identify and generate a correct path with them. For that, an auxiliar Perl script named renumCaIndex.pl is included in the downloaded set of files. The script reassigns the atom index numbers in the target structure according to the new ones given to the origin structure after the MD setup process. The script usage is very easy (see below). 

machine:/home/path> perl renumCaIndex.pl 
Usage: perl renumCaIndex.pl [origin] [target] [new]
Example: perl renumCaIndex.pl mdsystem.pdb frame_20.full.pdb targetTMD.pdb

Please note that we strongly recommend to run the structural alignment before reassigning atom indexes. That is because almost all software packages running structural alignments renumber the atom indexes from 1 to n (as it also happens with MD setup software). The fourth step once the structures are correctly prepared is the biased MD simulation using plumed. An input file example to run a Targeted MD simulation using plumed can be found in the tgz package downloaded (*.plumed.dat). This file contains the minimum commands to run a biased TMD from the selected origin structure to the target one. Before launching the simulation, the REFERENCE file name, corresponding to our new target structure, with the correct atom indexes and structurally aligned to the origin structure should be replaced in the example file (in the example below, "frame_5.plumed.aligned.pdb" is the name that should be replaced). 

# TransAtlas automatically generated plumed.dat file
# Targeted MD configuration
# creates a CV that measures the RMSD from a reference pdb structure
# the RMSD is measured after OPTIMAL alignment with the target structure
rmsd: RMSD REFERENCE=frame_5.plumed.aligned.pdb TYPE=OPTIMAL
# the movingrestraint
restraint: ...
        MOVINGRESTRAINT
        ARG=rmsd
        AT0=0.0 STEP0=0		KAPPA0=0
        AT1=0.0 STEP1=5000000	KAPPA1=10000
...
# monitor various restraint outputs
PRINT STRIDE=100 ARG=* FILE=TransAtlas.TMDplumed.out

The plumed configuration file can be further modified to track other Collective Variables apart from the RMSD. For example the CVs identified by TransAtlas analysis represented in the visualization section (see corresponding help section) can be added to this file. Please note that in this case the residue numbers shown for the CVs in TransAtlas should be transformed to the corresponding atom indexes so that plumed could correctly identify them (see an example in Targeted MD plumed help pages). A Gromacs execution is shown below, in order to illustrate how this plumed configuration file can be used with Gromacs MD package. Please note that plumed is not just coupled to Gromacs MD package, as it can be also used together with AMBER, NAMD, LAMMPS or QESPRESSO (see plumed compatible MD codes). 

gmx_mpi mdrun -v -s topol.tpr -o biasedTMD.ret.trr -c biasedTMD.ret.gro -e biasedTMD.ret.edr -g biasedTMD.ret.log -x biasedTMD.ret.xtc -plumed 1ba2_A_1urp_A.1_to_5.TMD.plumed.dat

This biased MD simulation will generate a plumed output file named TransAtlas.TMDplumed.out with information about the restraints applied and RMSd distance between origin and target throughout the simulation. If added, information about the CVs progress along the simulation could also be found in this output file. plumed out plot