PROFASI  Version 1.5
Restraining simulations with a torsional angle potential

Sometimes we are interested in keeping one of the proteins relatively fixed while allowing the other chains to move around freely. Sometimes we have a good idea about the secondary structure of a certain region of the chain, and want the simulations to not drift away from it too easily, while also not making that region entirely immobile. One way to guide the simulations in such cases is to use an energy penalty for backbone angles which differ from their preferred values. To use such a term in a simulation, you can write the following line in the settings file:

  force_field FF08DR=FF08+Extras:DihedralRestraints(restr_descr.xml)

where restr_descr.xml is an XML file describing the restraints (see below).

Equivalently, you can specify the use of the restraints in the command line.

  $ BasicMCRun --add_chain_pdb 1 monster.pdb::A -lcyc -1 -ncyc 100000 \
  --force_field "FF08DR=FF08+Extras:DihedralRestraints(restr_descr.xml)"

Notice that quotation marks are needed around the force_field option in the command line.

The above syntax can be used for any ProFASi simulation program.

Configuring distance restraints using an XML file

A dihedral restraints energy term must be configured so that it knows which angles to apply restraints on, and what the restraints look like. This is done using an XML configuration file. This is the format of the XML file:

In practice, it would be cumbersome to generate large XML files with such a nested structure. But ProFASi's XML module has a template parsing facility which can be used to do this much more elegantly. Let's say we have generated a long array of "desired" values for the dihedral angles with some notion about the uncertainity (width) associated with each angle. We have a file like this (let's say the file is called dh_con.table ):

::b:5 2.74191 0.5 10 1 1 2.74191 20
::b:6 -1.81689 0.5 10 1 1 -1.81689 20
::b:7 2.17992 0.5 10 1 1 2.17992 20
::b:8 -1.56731 0.5 10 1 1 -1.56731 20
::b:9 2.62498 0.5 10 1 1 2.62498 20
::b:10 -2.22879 0.5 10 1 1 -2.22879 20
::b:11 1.77849 0.5 10 1 1 1.77849 20
::b:12 -1.39277 0.5 10 1 1 -1.39277 20
::b:13 1.99491 0.5 10 1 1 1.99491 20
::b:14 -1.82772 0.5 10 1 1 -1.82772 20
::b:15 0.678933 0.5 10 1 1 0.678933 20
::b:16 -1.01753 0.5 10 1 1 -1.01753 20
::b:17 2.45219 0.5 10 1 1 2.45219 20
::b:18 -1.22173 0.5 10 1 1 -1.22173 20
::b:19 -0.486947 0.5 10 1 1 -0.486947 20
...

The format of the above file is entirely up to you. It does not matter which column you write the DOF identifier string, which column you put the mean etc. But for the discussion, let's say you want column 7 (counting from 1, like in awk) to be the mean, column 8 to be the kappa and column 4 to be the weight, and column 1 to be the DOF identifier string. You then make an XML file with the following contents (restr_descr.xml):

<dihedral_restraints>
  <formatted_data>
    <format name="restraint">
      <dof_id>$1</dof_id>
      <parameters>
        <mean>$7</mean>
        <kappa>$8</kappa>
        <weight>$4</weight>
      </parameters>
    </format>
    <import_data>dh_con.table</import_data>
  </formatted_data>
</dihedral_restraints>

ProFASi will see this XML file as follows:

<dihedral_restraints>
  <restraint>
    <dof_id>::b:5</dof_id>
    <parameters>
      <mean>2.74191</mean>
      <kappa>20</kappa>
      <weight>10</weight>
    </parameters>
  </restraint>
  <restraint>
    <dof_id>::b:6</dof_id>
    <parameters>
      <mean>-1.81689</mean>
      <kappa>20</kappa>
      <weight>10</weight>
    </parameters>
  </restraint>
  <restraint>
    <dof_id>::b:7</dof_id>
    <parameters>
      <mean>2.17992</mean>
      <kappa>20</kappa>
      <weight>10</weight>
    </parameters>
  </restraint>
  ...
</dihedral_restraints>

We did not specify a distribution type, so the von Mises or circular normal form will be assumed for every restraint. The parameters are as found in the parameters node.

See Also
Simulations with distance restraints

PROFASI: Protein Folding and Aggregation Simulator, Version 1.5
© (2005-2016) Anders Irbäck and Sandipan Mohanty
Documentation generated on Mon Jul 18 2016 using Doxygen version 1.8.2