pdb files in Rosetta format

7 posts / 0 new
pdb files in Rosetta format
#1

Hi,

I want to use rosetta to refine my NMR structure. Could someone tell me what is the rosetta format in pdb files, and how to add missing N- and C-term residues to my generated structure? Thank you very much.

Best regards,

Post Situation:
Mon, 2011-10-31 08:52
zhisheng

With some caveats, Rosetta should be able to use standard-formatted PDB files as-is. Non-standard protein residues may present some issues. Selenomethionine is the biggest offender (you need to change it to regular Met), but others may cause grief. Except for some metal ions, most non-protein ligands aren't recognized without additional effort. DNA works for the most part, but RNA requires special handling. The other big issue, especially for NMR people, is the hydrogen naming convention - if I recall correctly Rosetta follows the PDB's hydrogen naming conventions, which might not match those used by other structure producing tools. Note that if hydrogens are missing (or misnamed) Rosetta is able to rebuild them from ideal geometry. (It can also do that for missing side chain heavy atoms.)

As for adding missing residues, I'm not sure which program you're intending to use the input structure for, but if it is as a template in a homology modeling protocol, usually you don't need to have all of the residues in the template, as Rosetta can tell from the provided alignment that there are missing residues, and will insert and rebuild them to the best of its ability.

Mon, 2011-10-31 10:13
rmoretti

Regarding selenomethionine: It has been my experience that it works ok. There is code to catch MSE (selenomethionine) in the PDB reader:

 rosetta-3.3/rosetta_source/src/core> ack-grep MSE io/pdb/file_data.cc 370: else if ( name == "MSE" ) { 371: TR << "Reading MSE as MET!" << std::endl; 399: TR << "Reading Selenium SE from MSE as SD from MET" << std::endl;

Regarding adding terminal residues: Rosetta has no good way of doing it. If you can tell us why you want to do it, we can guide you towards the partial solutions that exist. There is a loop-length-size-change demo within the AnchoredDesign demo that may be able to handle termini changes as well (I didn't write it recently so I forget.)

Mon, 2011-10-31 11:02
smlewis

Hi,

Thank you very much for reply. I used chemical-shift rosetta to generate the structures. In order to get a better results I removed the N- and C-terminal residues. Now I want to add them and refine the structure. Thank you very much.

Cheers,

Tue, 2011-11-01 09:31
zhisheng

Good morning. Does it mean that Rosetta FF treats selenomethionine fully? If so, I would like to see the parameters chosen. The reason is that I would like to treat a system containing a protein rich of selenomethionine, first with Rosetta, then probably with charmm or amber FF. However, I was unable to see selenomethionine treated by any classical MD FF.
thanks, francesco pietra

Tue, 2012-01-31 23:48
chiendarret

Very sorry. I had not read smlewis' message with the due attention. I see that even Rosetta treats selenomethionine as methionine. francesco pietra

Tue, 2012-01-31 23:52
chiendarret

We consider the Se-MET = MET a feature, but if you wanted to define specific parameters for selenomethionine you could. It would be simplest to give it the "wrong" name, like "AAA" instead of "MSE" to prevent rosetta from auto-correcting it, then create selenomethionine as a custom residue type using the noncanonical residues tools (there are several threads on it).

Wed, 2012-02-01 06:59
smlewis