params file for beta-amino acids

Member Site › Forums › Rosetta 3 › Rosetta 3 – General › params file for beta-amino acids

Viewing 2 reply threads
  • Author
    Posts
    • March 30, 2021 at 1:30 am #3733
      Anonymous

        Hi all,

        I’ve been trying to insert a beta-amino acid into peptide, let’s say, beta-homo-leucine. It seems that a params file for beta-hLeu is needed. Unforutnately, there is no params file for beta-hLeu in database in Rosetta, and molfile_to_params.py only works for alpha-amino acid. I’ve also visited https://rosie.rosettacommons.org/documentation/beta_peptide_design, but did’t find helpful solution. So my question is:

        Is there any script for generating params file for beta-amino acid?

        Any helpful information is appreciated. Thank you!

         

      • March 30, 2021 at 3:17 am #15831
        Anonymous

          When you say beta-homo-leucine, do you mean “leucine, but there’s an extra backbone carbon atom” (which happens to be chemically equivalent to “homoleucine, but the nitrogen is bonded to the beta carbon”)? In that common case B3L is the parameter set you want.

          Separately, molfile_to_params_polymer.py definitely works for betas.

          • April 1, 2021 at 5:04 am #15839
            Anonymous

              Thank you for your reply! 

              The “beta-homo-leucine” I’m talking about is indeed the same as B3L in Rosetta database, and I was wrong to assume they’re different things just by the way they’re named.

              And to learn to generate params file for beta-aa, I took B3L as an example. Compared with B3L.params in Rosetta database, my result:

              1. got wrong CHIs(containing dihedral angles of backbone), so I deleted them manually. I wander if it’s ok to do that.

              2. had a different atom name for alpha-C, which in my result was CB1 but CM in Rosetta database. I don’t know how the atom tree is spanned in the latter and if it will matter.

               

              Anoter question is: I used FlexPepDock Refinement Protocol with a peptide containing B3L as input. But it got an error:

              ”’

              core.chemical.GlobalResidueTypeSet: Loading (but possibly not actually using) ‘B3L’ from the PDB components dictionary for residue type ‘pdb_B3L’

              core.conformation.Residue: [ ERROR ] New Residue cannot be aligned with template Residue, because the numbers of main-chain atoms differ.

              protocols.jd2.JobDistributor: [ ERROR ] 

              ”’

              I tried to substitute B3L with a natural amino acid and it worked. So I guessed this FlexPepDock Refinement Protocol is not compatible with beta-aa. How should I modify this protocol for beta-aa? Or change with other Dock Protocols?

              Thank you!

            • April 1, 2021 at 6:58 pm #15844
              Anonymous

                That might be a research question for flexpepdock! I assume they’re trying to make fragment moves and using alpha amino acid fragments. That’s going to be really hard.

                 

                I am tempted to say “write your own Monte Carlo protocol” but that is a tall order. What’s your scientific use case?

            • April 1, 2021 at 6:59 pm #15845
              Anonymous

                Oh and yes — atom naming is expected to be inconsistent with the raw molfile to params output. You do have to make some manual changes to echo the rosetta params style. At least it calculates all the internal coordinates right!

            • Author
              Posts
            Viewing 2 reply threads
            • You must be logged in to reply to this topic.