LIGPLOT of Val-Trp dipeptide in 3tmn		DIMPLOT of interface between chains A and B in 3sdp

1. Introduction

• LIGPLOT - automatically generates schematic diagrams of protein-ligand interactions for a given ligand in a PDB file.
• DIMPLOT - plots interactions across a selected protein-protein or domain-domain interface.

2. Paths and directories

Edit the text fields as follows:

PDB paths. PDB files are most conveniently identified simply by the 4-character PDB code or identifier. If your system contains one or more locations where PDB files are stored, you can define a "template" path which will allow LIGPLOT⁺ to find the relevant PDB file from its 4-character code alone. The files can be gzipped or not.

Alternatively, if you are connected to the internet when you run LIGPLOT⁺, the required file can be retrieved by ftp from a given ftp site.

The 4 characters of the PDB code in each template are represented as abcd. These appear in the template within square brackets.

For example, the template

C:/roman/pdbsum/pdb/pdb[abcd].ent

C:/roman/pdbsum/pdb/pdb1ral.ent

ftp://ftp.ebi.ac.uk/pub/databases/rcsb/pdb-remediated/data/structures/divided/pdb/[bc]/pdb[abcd].ent.gz

ftp://ftp.ebi.ac.uk/pub/databases/rcsb/pdb-remediated/data/structures/divided/pdb/ra/pdb1ral.ent.gz

The above ftp address retrieves gzipped PDB files from the PDBe ftp site at the EBI. A full list of possible ftp addresses is given below:

PDBe		ftp://ftp.ebi.ac.uk/pub/databases/rcsb/pdb-remediated/data/structures/divided/pdb/[bc]/pdb[abcd].ent.gz
PDBj		ftp://pdb.protein.osaka-u.ac.jp/v3/pub/pdb/data/structures/divided/pdb/[bc]/pdb[abcd].ent.gz
RCSB PDB		ftp://ftp.wwpdb.org/pub/pdb/data/structures/divided/pdb/[bc]/pdb[abcd].ent.gz

The Het Group Dictionary. LIGPLOT⁺ gets its information about ligand molecules from the Het Group Dictionary. This defines the atom names, connectivities and bond orders of every Het Group in the PDB. You can download a copy of this dictionary from the wwPDB at:

ftp://ftp.wwpdb.org/pub/pdb/data/monomers/components.cif

It is a good idea to download it regularly to always have an up-to-date copy.

Record the full path and filename of the Het Group Dictionary here so that the programs know where to find it.

Temporary directory. LIGPLOT⁺ needs a directory where it can puts its temporary working files (which it deletes once it has done with them). Use this parameter to define a directory to which you have write-access and which LIGPLOT⁺ can use for its working files.

Name of RasMol executable. Enter the full path and name of the RasMol executable file. This will allow you to display the 3D coords of the given LIGPLOT/DIMPLOT diagrams. If you do not have RasMol, enter NONE here, although it is recommended that you download and install a copy of the program.

Linux users might find it more convenient to enter the RasMol path using the following format:

xterm -e rasmol -script

Save. Press the Save button to store your paths. If you should need to alter the paths at some leter date, you can do so via the Program paths menu item in the Edit menu.

3. Generating a LIGPLOT/DIMPLOT diagram

Alternatively, you can browse for a PDB file in your local directory system by clicking the Browse button.

When the file has been loaded, you will see a summary of the ligands and metals in the PDB file, if any, as shown below for PDB entry 1a95.

Clicking on the LIGPLOT or DIMPLOT tabs allows you to choose which of the two programs to run.

• LIGPLOT - Select which ligand you wish to plot from the list, or enter your own residue range in the box. The residue range can be one of the following:

  Example		Description
  18 20 A		Residue range: start-residue-no. end-residue-no. chain-id If the chain-id is blank it should be omitted
  NAG 18 MAN 20 A		Residue names and numbers, plus chain id.
  -n818 108 -n818 108 A		Where the residue name is numeric, prefix it with "-n" to indicate that it is the residue name rather than the residue number.
  "MTE" 1 "  G" 2 B		Where a residue name includes leading blanks (eg as in the old definition of the nucleic acid bases, A, C, G and T), enclose it in quotes and include the spaces.
  "-nMTE" 1 "-n  G" 2 B		An alternative for the case above.

• DIMPLOT - Select which protein chain pair you would like plotted, or enter the residue ranges of the chains or domains you want to plot. Enter each residue range as: N1-N2 C, where N1 is the start residue, N2 is the end residue, and C is the chain id. If a domain is split, you can define it using several residue ranges, separating each range with an ampersand. For example

  1-136 A & 338-371 A

When the plot has been generated it will appear on screen and can now be edited interactively, as described next.

4. Editing the plot

a. Moving residues, atoms, and text items

Note, that to switch between moving whole residues and single atoms click the Move residue/move atom button at the bottom-right of the frame to toggle between the two modes.

Note also that the hydrophobic contact groups cannot be moved in the Move atom mode.

b. Panning and zooming

Use the right mouse button to click-and-drag on any blank area of the plot to zoom in and out of the plot.

c. Recentre

d. Rotating residues

To rotate about this atom, click-and-drag any other atom in the residue. Release the button when you have reached the desired position.

To deselect the atom, right- or left-click on any blank part of the plot, or on any other residue.

e. Flipping about a bond

Firstly, select the bond by clicking on it with the right mouse button. A marker will appear over the selected bond.

To flip about this bond, click with the left mouse-button on any atom on the side of the residue to be flipped. A single click will perform the flip. Clicking a second time on any of the flipped atoms will reverse the flip.

To "unselect" the flip-bond, click with the right or middle mouse button on any blank area of the plot. The marker will disappear.

f. Editing text labels

Select the text item by right-clicking on it. A box containing the text string will appear. You can edit the text as required. Click the OK button to accept the new version, or the Cancel button to retain the previous text label.

5. Multiple plots

The first plot is generated in the usual way. Each subsequent plot is fitted to the first by a sequence-based comparison of the two binding sites. This identifies equivalent residues in the 3D structures and these equivalences drive the generation of the second LIGPLOT. If the program is unable to match the binding sites based on their sequences, it tries to fit the ligands using a graph matching procedure. For very distantly related proteins it is better to use a structural alignment. Section 7 describes how to import such an alignment.

a. Loading additional plots

The new plot, when generated, will become the foreground plot; any current plot(s) will go into the background.

The left-hand example above shows a plot of the interactions between guanine 600 in PDB entry 2pwu and the protein residues. The right-hand plot shows superposed on it a plot of a similar molecule () bound to PDB entry 1q2r. The red circles and ellipses identify the residue on the latter plot that are equivalent to the underlying residues from the first plot.

LigPlot⁺ attempts, as best it can, to place residues in the new plot on top of the equivalent residues in the old one. It does this by first performing a simple sequence alignment between the set of interacting residues from each structure. The equivalenced residues in this alignment are then used to superpose the 3D coordinates from both PDB files. This may throw up more pairs of equivalenced residues, being those that significantly overlap in the superposition. The residue equivalences are then used to "drive" the generation of the new LIGPLOT; the 2D locations from the first plot restrain the positions of the corresponding residues in the second plot.

The superposition may not always be successful if either the two proteins are quite different, or the second ligand is very large relative to the first.

The "Split screen" button at the bottom of the frame separates the two (or more) plots, and arranges them in the window.

Two plots, as in the above example, are shown side by side:

The "Merge back" button will restore the plots to overlap mode.

b. Selecting the active plot

c. Viewing structures in RasMol

Clicking this will pop up a RasMol window containing the superposed ligands, in 3D, with H-bonds added in cyan and atoms involved in non-bonded contacts represented by dot surfaces.

d. Switching off the background plots

Then, deselect the Show inactive plots option, circled in red below.

6. Plot parameters

The colour options differ slightly depending on whether your plot is a LIGPLOT or a DIMPLOT. For the former, bonds and atoms are classified as ligand or non-ligand. For the latter, the two classes are: interface 1 and interface 2.

If you have two or more plots overlaid, you will be asked, after making your changes, whether the new parameters are to apply to the currently selected plot only, or to all plots. Thus you can, say, have the bonds or labels shown in a different colour in each plot. The background parameters always apply to all plots.

a. On/off selection

The items are grouped into: Labels, Atoms, Bonds and Miscellaneous.

b. Sizes selection

The items are grouped into: Label sizes, Atom and residue sizes and Bond widths.

The left-hand panel above shows the LIGPLOT options while the right-hand panel contains the DIMPLOT options.

c. Colour selection

Background, Labels, Atoms and Bonds.

The left-hand panel above shows the LIGPLOT options while the right-hand panel contains the DIMPLOT options.

Note that, for ligand and non-ligand bonds there is a special extra colour labelled "ATOM". If this "colour" is selected, the bonds will be coloured such that each half is of the colour of the atom bonded at that end.

7. Adding missing H-bonds

If you need your plot to have an H-bond that HBPLUS has missed, you can add it as follows:

1. Edit your PDB file and add the missing H-bonds, or non-bonded contacts, as "HHB" and "NNB" records, respectively. These records should come before the ATOM records in the file.

   The format of each "HHB" or "NNB" record should be formatted exactly as shown below, where the dots represent blank spaces.

   Key   <----Atom 1 --->     <----Atom 2 --->    Dist
   HHB...RRR.C.NNNNI.AAAA.....RRR.C.NNNNI.AAAA....DDDD
   

   where		HHB		defines the record type: HHB for H-bonds, NNB for non-bonded contacts
   		RRR		is the 3-character residue name
   		C		is chain identifier (which may be blank)
   		NNNN		is the right-justified residue number (1-9999)
   		I		is the insertion code, or blank if none
   		AAAA		is the 4-character atom name in standard PDB format
   		DDDD		is the distance between the two atoms (to 2 places of decimals)

   Some example lines are given below:

   HHB...RRR.C.NNNNI.AAAA.....RRR.C.NNNNI.AAAA....DDDD
   HHB   HIS A  231   N       ASP A  226   OD2    2.77
   HHB   HIS A  231   ND1     ASP A  226   OD1    2.76
   HHB   HIS A  231   NE2     THR A 1317   O      2.80
   HHB   THR A 1317   N       ALA A  113   O      2.72
   HHB   ALA A  113   N       PHQ A  317   O2     3.31
   

2. After editing any required H-bonds or non-bonded contacts into your PDB file, rerun LigPlot⁺ on this new PDB file.

8. Removing unwanted H-bonds

So, for example, to remove the H-bond between the NH₂ of Arg286(A) and the O7 of your ligand, LIG1(B), you would add the following line to your PDB file:

-HB...RRR.C.NNNNI.AAAA.....RRR.C.NNNNI.AAAA....DDDD
-HB   ARG A  286   NH2     LIG B    1   O7     0.00

9. Plots from structural alignments

One way to help the program is to supply it with a structural alignment between the proteins, letting it know which residues are equivalent in 3D.

At present, LigPlot⁺ only accepts structural alignments in CORA format (ie from the CATH team's CORA program). It is planned to add additional formats in the future.

Import the alignment

The CORA file needs to be CORA-format v.1.1. The first line of the file will give the version number. Import using File-Import from the menu bar. Then browse to the appropriate CORA file and Open it.

A pop-up window will show you the list of PDB codes in the file and allow you to select which one you want to plot first.

After the first plot has been generated, you can add others on top of it via either File-Add or File-Open-PDB file.