Hi Matt, Waters refine OK (at least in my hands) as long as you stick to this format: ATOM 31567 O HOH W2244 15.221 -55.194 88.452 0.72 48.20 W O However, I'd be interested to know if the real problem is not your ligand(s) after all - because I am having exactly the same issue you described when I try to refine some hydrated ions... Ciao Luca On 15 Mar 2007, at 15:15, Matthew Bowler wrote:

Thanks for replies so far....

I removed all ligands and then waters - the waters seem to be the problem as the protein model is now refining, do they need to be treated/marked in a different way for phenix?

Thanks again, Matt.

On 15 Mar 2007, at 13:32, Nicholas Noinaj wrote:

...

matt,

can't offer much help, but have you tried to remove the ligand and try refinement? it would let you know if the pdb file or the presence of the ligand is the problem. at least let you would know where you should focus your attention.

if all else fails, have you tried getting parameters for your ligand from the PRODRG Server and using that for inputing the ligand parameters into phenix.refine directly when you setup the refinement job? (I think i have done this before???)

here is the website in case you aren't familiar... http://davapc1.bioch.dundee.ac.uk/programs/prodrg/

Again, Best of luck!

cheers, nick

-----Original Message----- From: Matthew Bowler To: [email protected] Date: Thu, 15 Mar 2007 13:00:42 +0000 Subject: [phenixbb] phenix.refine problems with my pdb

Dear All, I am trying to run phenix.refine but I keep getting the message:

Sorry: Fatal problems interpreting PDB file: Number of atoms with unknown nonbonded energy type symbols: 308 Please edit the PDB file to resolve the problems and/or supply a CIF file with matching restraint definitions. Note that elbow.builder is available to create restraint definitions.

I have run my ligands through elbow and used the output .cif but I still get the same message. I can find no info on this problem in the manual or the BB, can anyone help? Thanks in advance, yours, Matt.

Matthew Bowler MRC Dunn Human Nutrition Unit Wellcome Trust / MRC Building Hills Road Cambridge CB2 2XY Tel: 0044 (0) 1223 252826 Fax: 0044 (0) 1223 252825

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Matthew Bowler MRC Dunn Human Nutrition Unit Wellcome Trust / MRC Building Hills Road Cambridge CB2 2XY Tel: 0044 (0) 1223 252826 Fax: 0044 (0) 1223 252825

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Hi Everybody, if you run phenix.refine like: % phenix.refine data.mtz model.pdb and it stops with something like: Sorry: Fatal problems interpreting PDB file: ... this means there is unknown (to Monomer Library) molecule in you file. The standard way to fix it is to get a cif (parameters file) by running eLBOW (which is a part of PHENIX and CCI Apps): % elbow.builder model.pdb --do-all In this case eLBOW will go though your PDB file and generate a cif file for all unknown ligands. The output file will contain all in its name, like: elbow.XXX.all.cif. So next time you run phenix.refine as: % phenix.refine data.mtz model.pdb elbow.XXX.all.cif. Also, in your case I'm really interested why water molecules in your file were causing the problem. I would really appreciate if you send me a couple of lines from your PDB file containing problematic waters, so I fix it and it will never happens again. Yes, you can also remove all waters and ask phenix.refine to build them anew; it does a good job on doing so: % phenix.refine data.mtz model.pdb ordered_solvent=true main.number_of_macro_cycles=5 Please let us know if you still have any questions / problems. Pavel. Matthew Bowler wrote:

Thanks for replies so far....

I removed all ligands and then waters - the waters seem to be the problem as the protein model is now refining, do they need to be treated/marked in a different way for phenix?

Thanks again, Matt.

On 15 Mar 2007, at 13:32, Nicholas Noinaj wrote:

...

matt,

can't offer much help, but have you tried to remove the ligand and try refinement? it would let you know if the pdb file or the presence of the ligand is the problem. at least let you would know where you should focus your attention.

if all else fails, have you tried getting parameters for your ligand from the PRODRG Server and using that for inputing the ligand parameters into phenix.refine directly when you setup the refinement job? (I think i have done this before???)

here is the website in case you aren't familiar... http://davapc1.bioch.dundee.ac.uk/programs/prodrg/

Again, Best of luck!

cheers, nick

-----Original Message----- From: Matthew Bowler To: [email protected] Date: Thu, 15 Mar 2007 13:00:42 +0000 Subject: [phenixbb] phenix.refine problems with my pdb

Dear All, I am trying to run phenix.refine but I keep getting the message:

Sorry: Fatal problems interpreting PDB file: Number of atoms with unknown nonbonded energy type symbols: 308 Please edit the PDB file to resolve the problems and/or supply a CIF file with matching restraint definitions. Note that elbow.builder is available to create restraint definitions.

I have run my ligands through elbow and used the output .cif but I still get the same message. I can find no info on this problem in the manual or the BB, can anyone help? Thanks in advance, yours, Matt.

Matthew Bowler MRC Dunn Human Nutrition Unit Wellcome Trust / MRC Building Hills Road Cambridge CB2 2XY Tel: 0044 (0) 1223 252826 Fax: 0044 (0) 1223 252825

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Matthew Bowler MRC Dunn Human Nutrition Unit Wellcome Trust / MRC Building Hills Road Cambridge CB2 2XY Tel: 0044 (0) 1223 252826 Fax: 0044 (0) 1223 252825

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

I have seen problems with tye PRODRG cif files as well, mainly due to the presence of 'inconsistent' torsion restraints. If you use elbow.builder , this should not be a problem (and comes with the phenix suite btw) You can feed elbow a smiles string, or (if you don't have that) give it a PDB file. If hydrogens or connect records are there, it will do a good job figuring out what the geometry is/should be. Cheers Peter ----- Original Message ----- From: Lucas Bleicher Date: Thursday, March 15, 2007 7:44 am Subject: Re: [phenixbb] phenix.refine problem CIF generated by PRODRG To: [email protected], PHENIX user mailing list

--- Nicholas Noinaj escreveu:

...

if all else fails, have you tried getting parameters for your ligand from the PRODRG Server and using that for inputing the ligand parameters into phenix.refine directly when you setup the refinement job? (I think i have done this before???) here is the website in case you aren't familiar... http://davapc1.bioch.dundee.ac.uk/programs/prodrg/

The message was not directed to me, but as PRODRG was cited, I'd like to comment on that.

I have been using the PRODRG server to generate parameters for my ligands for some time. When I use the CIF file on refmac it works perfectly. However, if I use the same file on phenix.refine the ligand becomes very distorted. I could do all protein refinement on phenix.refine and then run some cycles of refinement on refmac to do ligand refinement, but I'd really like to do all the refinement in phenix, which, for my cases, seems to produce much better results than refmac. Does anybody faced a problem like this? If so, did you succeeded in correcting it?

Lucas Bleicher

__________________________________________________ Fale com seus amigos de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com/ _______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb