Dealing with elemental ions

HADDOCK
1

The HADDOCK category is meant to discuss any HADDOCK-related issue. For general information about HADDOCK refer to http://www.bonvinlab.org/software/haddock2.2

2

The way to define ions for HADDOCK is explained in Box 3 of our Nature Protocols 2010 server paper:
The HADDOCK web server for data-driven biomolecular docking. Nature Protocols, 5, 883-897 (2010). Download the final author version here.

In short:

  1. the residue name should contain the absolute charge state
  2. the atom name should have the exact charge

E.g. for a zinc atom:

  • residue name: ZN2
  • atom name: ZN+2 (atom name with four characters should be shifted by one column to the left)
1 Like
3

Dear all,

I was hoping for some help in clarifying an issue:

I am using the webserver to dock two proteins, one of which has tetraluminium fluoride in it. This causes an error, since aluminium (AL) is an unsupported atom type. Does anyone have any advice on the best way to solve this?

Also, I’m not quite sure how to write the negative aluminium charge in the PDB, since the charge and residue name will not fit into the designated PDB column. See below for the example:

HETATM 2744 MG+2 MG2 C 512 187.954 211.305 314.482 1.00 7.62 Mg
HETATM 2745 AL-1 ALF C 513 190.924 211.597 312.551 1.00 23.61 Al
HETATM 2746 F1 ALF C 513 191.192 210.698 314.047 1.00 23.61 F
HETATM 2747 F2 ALF C 513 190.943 212.674 311.117 1.00 23.61 F
HETATM 2748 F3 ALF C 513 192.613 211.127 312.117 1.00 23.61 F
HETATM 2749 F4 ALF C 513 189.435 212.364 313.172 1.00 23.61 F

Any help would be gratefully received!

Cheers,
Alex

4

Well, first of all Aluminium is a positive ion…. +3 - so proper naming would thus be: AL+3 and AL3 as residue name

The main question is however: do you really need it for your docking? Or is it only some crystallisation buffer present in your PDB?

5

In this case, Aluminium is covalently bound to the four Fluorine atoms, giving it a charge of -1, hence the residue name is ALF (aluminium fluoride) rather than AL. The PDB I have is a homology model, and the aluminium fluoride is part of a ligand; ADP-aluminium fluoride (this mimics ATP hydrolysis, with the aluminium fluoride imitiating the orthophosphate). Apologies for not clarifying this.

Perhaps it’s not strictly necessary to include it, since I’m not allowing any of the interacting residues in semi-flexible refinement. Do you have any advice on whether it is necessary?

Thanks for the help

6

Then don’t specify any charges for it and see if it works…

Prof. A.M.J.J. Bonvin | Faculty of Science
Utrecht University | bonvinlab.orghttp://bonvinlab.org
a.m.j.j.bonvin@uu.nlmailto:a.m.j.j.bonvin@uu.nl | T: +31 30 2533859

7

In the end, the best solution was to separate all the fluorine and aluminium atoms, and just give them distance restraints to keep them all together.

Thanks for your help.