I'm looking for some clarification of how the Robetta server, and specifically the new TrRosetta algorithm, deals with structure prediction for membrane proteins.
When I submit a job I can see that Robetta is performing transmembrane helix prediction using the TMHMM server, so I assume that some kind of membrane modelling is taking place. However, when I tried TrRosetta on some small membrane proteins of approximately 100 amino acids, I find that polar residues are predicted to be exposed rather than buried as would be expected in a membrane. I have seen success for some larger membrane proteins though which gave sensible outputs.
Please redirect me if this is posted in the wrong forum.