pathtools
This is part of the mapping module
It is only available if you configure PLUMED with ./configure –enable-modules=mapping . Furthermore, this feature is still being developed so take care when using it and report any problems on the mailing list.

pathtools can be used to construct paths from pdb data

The path CVs in PLUMED are curvilinear coordinates through a high dimensional vector space. Enhanced sampling calculations are often run using the progress along the paths and the distance from the path as CVs as this provides a convenient way of defining a reaction coordinate for a complicated process. This method is explained in the documentation for PATH.

The path itself is an ordered set of equally-spaced, high-dimensional frames the way in which these frames should be constructed will depend on the problem in hand. In other words, you will need to understand the reaction you wish to study in order to select a sensible set of frames to use in your path CV. This tool provides two methods that may be useful when it comes to constructing paths; namely:

  • A tool that takes in an initial guess path in which the frames are not equally spaced. This tool adjusts the positions of the frames in order to make them equally spaced so that they can be used as the basis for a path CV.
  • A tool that takes two frames as input and that allows you to return a linear path connecting these two frames. The output from this method may be useful as an initial guess path. It is arguable that a linear path rather defeats the purpose of the path CV method, however, as the whole purpose is to be able to define non-linear paths.

Notice that you can use these two methods and take advantage of all the ways of measuring Distances from reference configurations that are available within PLUMED. The way you do this with each of these tools described above is explained in the example below.

Examples

The example below shows how you can take a set of unequally spaced frames from a pdb file named in_path.pdb and use pathtools to make them equally spaced so that they can be used as the basis for a path CV. The file containing this final path is named final_path.pdb.

plumed pathtools --path in_path.pdb --metric EUCLIDEAN --out final_path.pdb

The example below shows how can create an initial linear path connecting the two pdb frames in start.pdb and end.pdb. In this case the path output to path.pdb will consist of 6 frames: the initial and final frames that were contained in start.pdb and end.pdb as well as four equally spaced frames along the vector connecting start.pdb to end.pdb.

plumed pathtools --start start.pdb --end end.pdb --nframes 4 --metric OPTIMAL --out path.pdb

Often the idea with path collective variables is to create a path connecting some initial state A to some final state B. You would in this case have representative configurations from your A and B states defined in the input files to pathtools that we have called start.pdb and end.pdb in the example above. Furthermore, it may be useful to have a few frames before your start frame and after your end frame. You can use path tools to create these extended paths as shown below. In this case the final path would now consist of 8 frames. Four of these frames would lie on the vector connecting state A to state B, there would be one frame each at start.pdb and end.pdb as well as one frame just before start.pdb and one frame just after end.pdb. All these frames would be equally spaced.

plumed pathtools --start start.pdb --end end.pdb --nframes 4 --metric OPTIMAL --out path.pdb --nframes-before-start 2 --nframes-after-end 2

Notice also that when you re-parameterize paths you must choose two frames to fix. Generally you chose to fix the states that are representative of your states A and B. By default pathtools will fix the first and last frames. You can, however, change the states to fix by taking advantage of the fixed flag as shown below.

plumed pathtools --path inpath.pdb --metric EUCLIDEAN --out outpath.pdb --fixed 2,12
Glossary of keywords and components
The atoms involved can be specified using
--start a pdb file that contains the structure for the initial frame of your path. For more information on how to specify lists of atoms see Groups and Virtual Atoms
--end a pdb file that contains the structure for the final frame of your path. For more information on how to specify lists of atoms see Groups and Virtual Atoms
Or alternatively by using
--path a pdb file that contains an initial path in which the frames are not equally spaced
Compulsory keywords
--fixed ( default=0 ) the frames to fix when constructing the path using –path
--metric the measure to use to calculate the distance between frames
--out the name of the file on which to output your path
--arg-fmt ( default=f ) the format to use for argument values in your frames
--tolerance ( default=1E-4 ) the tolerance to use for the algorithm that is used to re-parameterize the path
--nframes-before-start ( default=1 ) the number of frames to include in the path before the first frame
--nframes ( default=1 ) the number of frames between the start and end frames in your path
--nframes-after-end ( default=1 ) the number of frames to put after the last frame of your path
Options
--help/-h

( default=off ) print this help

--arg the arguments that should be read in from the pdb files