Creating Index Files

Introduction

Sometimes we are interested in only analyzing specific parts of a protein in a simulation. GROMACS index files (NDX) contain information on atom groups such as protein-associated atoms, ions, and other molecules (e.g., ATP or ADP). Index files are used with the -n option in various GROMACS modules to analyze portions of your protein.

By default, GROMACS will use a preset of groups if no index file is supplied.

We can make custom index files for special groups through the gmx make_ndx command.

You should have the following files:

• Topology file (.tpr)

Using gmx make_ndx

To create an index file:

gmx make_ndx -f topology.tpr -o output.ndx

You can additionally supply a pre-existing index file to modify using the -n option.

In either case, you will be presented with the index editor containing default atom groups, grouping options, and a prompt to enter commands (>).

Entering q at this point will save a NDX file containing only default groups.

Navigating the index editor

To create special groups, we will need to use the grouping options. Type h to show the help menu.

The help menu explains how to use each grouping option. It also provides a few examples at the bottom showing how to use multiple options with operators (! & |).

Hitting ENTER with an empty command prompt will repopulate the group list and bring you to the original menu.

---

Generally, you will not need to do overly complicated index file manipulation.

To select a range of atoms, like the first 100:

a 1-100

Similarly, for the first 100 residues:

r 1-100

---

You can also mash commands together using operators (! & |).

To select the backbone and side-chain hydrogen atoms of residues 300 to 350, but not residue 325:

4 | 9 & r 300-350 & ! r 325

Breaking this command down:

• 4 | 9 selects all atoms that are part of either groups 4 OR (|) 9.
  • Using AND (&) here will return an empty group, as & looks for atoms shared by both groups.
• & r 300-350 & ! r 325 selects residues 300 to 350 and NOT (& !) residue 325.

Note that if your protein has multiple chains, r <residue number> will select all residues with that number. To select residues in specific chains only, you will need to create chain-only groups first.

Splitting protein chains

A common way of using gmx make_ndx is to create special groups for protein chains. The splitch <group number> option splits the protein into chains that contain only atoms from the specified group.

In the group list above, the entire protein is defined in Group 1 and the backbone is defined in Group 4.

splitch 1 will create individual groups for each protein chain. These groups will contain all atoms of that chain, including those in side-chains.

splitch 4 will also create individual groups per chain, but these groups will only contain the atoms of the protein backbone.

This particular protein has six chains. After repopulating the group list with ENTER, we see both grouping options created six groups each, with splitch 1 groups boxed in red (21 to 26) and splitch 4 boxed in yellow (27 to 32).

Note

Splitting actions like splitch, splitat, and splitres cannot be used with other options or operators. Instead, these must be executed sequentially.

Additional Resources

• gmx make_ndx