Unsolved

Hi!!

I had been suscesfully using ROSIE to generate Fab models for a while and lately the server has not able to recognize the CDRL3 and CDRH3 in my antibody sequences. I also tried a sequence that was previously used and modeled but now the server gave an error  

ERROR: CDR L3 cannot be recognized !!!  L3 pattern: C[A-Z]{1,15}(L|F|V|S)G[A-Z](G|Y)

and I dont undestand why  the CDR cant not be recognized now. The pattern of the input seqeunces looks correct. 

I would really aprreacite some help in this matter

 Hey,

I am using Rosetta to build my protein from Ca-trace to all-atom, and then to refine this structure against cryo-em density.

However, when I use these two flags:- RBSegmentRelax::cst_wt 1.0  and -RBSegmentRelax::cst_width 4.0, I am getting the error:

Option matching -RBSegmentRelax:cst_wt not found in command line top-level context. 

When I remove these two flags, everytging works perfectly.

So, my question is: do these two flags are still used by ca_to_allatom application?

Dear structure predictors,

For extracting an alignment with the best template I used HHPred.

Which than predicted a chain of a PDB which is not the best one, because it doesn´t contain an important N-terminal part of my target. 

I would like to use another chain of this PDB but HHPred doesn´t predict it.

Is there a possibility to have a sequence structure alignment between a defined PDB and my target protein sequence with same quality HHPred does?