You are here


2 posts / 0 new
Last post


I've been trying to create a pose with trans-ACPC, but when I use the X[TAC] in the sequence, it doesn't seem to know what residue it is. I know there's a params file for it... I tried loading a pdb structure with ACPC in it as my pose to see what'd happen and it says that:

"ERROR: Error in core::scoring::methods::RamaPreProEnergy::residue_pair_energy(): The RamaPrePro term is incompatible with cyclic dipeptides (as is most of the rest of Rosetta)."

Is it possible to design beta-peptides in PyRosetta? Am I doing something wrong?

Thank you!



Post Situation: 
Wed, 2019-12-04 03:23

Try "X[transACPC]".  Three caveats, though:

1.  If you're building a sequence from scratch (e.g. using the PeptideStubMover), this should work.  If you're trying to mutate a chain of glycines to a sequence with the packer or something similar, this will fail.  You can't introduce a mutation that changes backbone geometry -- Rosetta doesn't know where to put the extra carbon atom (or rather, Rosetta doesn't know how to update the structure given that the rigid-body transform from the residue's lower polymeric connection to its upper polymeric connection has changed).  You can only mutate alpha-amino acids to alpha-amino acids and beta-amino acids to beta-amino acids.

2.  The pro_close term will not work for transACPC.  Try replacing this with the ring_close term in the scorefunction that you're using.  I _think_ that works with transACPC, but I can't remember.  There is currenly no mainchain potential for transACPC, which will affect accuracy of many protocols.  (That is, you can build a model containing this resiude, but don't rely on the energies that you calculate for anything important.)

3.  I added the transACPC params file a very long time ago, and haven't touched it in some time.  It may or may not be up to date and functional with current Rosetta.  Please use at your own risk.

Wed, 2019-12-04 12:56