[buster-discuss] TNT and FE-S clusters!

[Pietro Roversi]

Dear Rams,
             here is how I would do it:

1. Prepare a PDB file with your Fe-S cluster only. Ideally, this should come from an idealised structure like the ones that the CORINA server produces (see ) but an good existing PDB entry is also a good source I guess.

2. Run Gerard Kleywegt Uppsala's xplo2d program (it is part of the x_util suite) using your sulphur cluster PDB file as input. In the following I assume the cluster atoms are in a residue named FES in that PDB file. If you ask for it, this will generate a TNT dictionary for the cluster. 

Save this dictionary file in your Buster-TNT projects datafiles directory, say call it FeS.cluster.tnt.dat. Notice that the latest version of xplo2d will add the BCORR cards that restrain the Bfactors for each pair of bonded atoms, but restraining the DeltaB to be zero plus or minus 10 (5?) so if you have a better idea about the DeltaBs edit those cards. Obviously if you are not running the latest xplo2d binary you will have to edit the file and add to the dictionary one BCORR card for each BOND card, saying something like:

GEOMETRY FES BCORR 0.0 10.0 FE1 S1
GEOMETRY FES BCORR 0.0 10.0 FE1 S2 
....
GEOMETRY FES BCORR 0.0 10.0 FE1 SN
....
 
3. I assume you have your TNT sequence file, for example having run $BDG_home/bin/buster/pir2seq.pl starting from a 1-letter sequence file in pir format, or however you produced it.

To that sequence file add one entry for each independent copy of the cluster you have in the PDB (say you have three clusters with residue name FES and residue numbers 1001, 2001 and 3001 in chain A:

RESIDUE FES A|1001
RESIDUE FES A|2001
RESIDUE FES A|3001

This will make sure that the geometry of the FES residue as per the geometry file is enforced. If the clusters are in chain W on their own do not forget to declare that chain at the top of the sequence file.

4. Add the cluster dictionary to the dictinaries used by TNT in the appropriate box in the input interface:

$BDG_datafiles/FeS.cluster.tnt.dat

I assume that the clusters are not covalently bound to the protein so you don't need to specify any protein-cluster bonds nor declare them in the sequence file.

I hope this helps! but do not hesitate to write again if you still cannot get things to work.

Ciao

Pietro