Using D-amino acids in Rosetta docking

The unknown atom name error is not occuring when Rosetta is reading in your PDB file, it is occuring when other code in the program is trying to access the H atom on that residue. Proline doesn't have a amide hydrogen, so trying to look it up is a pretty legitimate failure.

I think in both cases these failures are likely caused somewhere in the scorefunction. The standard scorefunctions make many assumptions about having normal protein; the scorefunctions used for noncanonicals (and Ds) do not. I've emailed the noncanonicals guy for advice on which scorefunction to use.

The NCAA and D-CAA code was added by me. Hopefully I will be able to answer all your questions.

• D-Proline Problem:
  The problem with the D-proline occurs when the base residue type parameter files are patched to create the different variants such as the N-terminus and C-terminus. For example, when the N-term patch (rosetta_database/chemical/residue_type_sets/fa_standard/patches/NtermProteinFull.txt) is applied to the set of residue types it removes the hydrogen (H) attached to the backbone nitrogen (N) and then adds three hydrogens (1H, 2H, 3H) and sets up the bonds, internal coordinates, and properties for the new atoms and new residue type (each applied patch creates a new residue type). If you look at the above file you will see that there is a special case for glycine and proline. Since the base proline doesn't have the H to remove it starts by adding two hydrogens (1H, 2H) and then setting up their bonds, internal coordinates, and properties. The segfault is occurring because D-PRO also lacks that H but since there is not a special case for D-PRO, the general case is being used instead. When Rosetta tries to delete the non-existant H from D-PRO it crashes because it can't find it. There is a similar problem with the N_acetylated.txt, ShoveBB.txt, VirtualBB.txt, VirtualNterm.txt, protein_cutpoint_upper.txt.

  Modified patch files that include a special case for DPRO can be found at the links bellow...
  http://carl.bio.nyu.edu/~renfrew/for_moehle/NtermProteinFull.txt
  http://carl.bio.nyu.edu/~renfrew/for_moehle/N_acetylated.txt
  http://carl.bio.nyu.edu/~renfrew/for_moehle/ShoveBB.txt
  http://carl.bio.nyu.edu/~renfrew/for_moehle/VirtualBB.txt
  http://carl.bio.nyu.edu/~renfrew/for_moehle/VirtualNterm.txt
  http://carl.bio.nyu.edu/~renfrew/for_moehle/protein_cutpoint_upper.txt

• D-Arginine Problem (really two problems but you had not encountered the second yet):
  • The noncanonical amino acid (NCAA) rotamer libraries are not in the release because they are rarely used and quite large (~450MB). The DARG and DPRO rotlibs can be found at the links bellow...
    http://carl.bio.nyu.edu/~renfrew/for_moehle/dpro.rotlib
    http://carl.bio.nyu.edu/~renfrew/for_moehle/darg.rotlib
  • Steven correctly anticipated that there would be a scoring issue using the the standard Rosetta scoring functions with the D-amino acids. If you would like to use the D-amino acids you will need to use the MM_STD scoring function by adding "-score::weights mm_std". This energy function has had the knowledge-based terms removed and replaced with equivalent physically-based molecular-mechanics potentials. There may still be problems if certain rosetta protocols ignore that command line option and hard code which energy methods they use. There is also no centroid scoring function for the NCAAs so all of your docking will have to be done in fullatom mode.