According to my understanding, Rosetta does not consider/design disulfide in simulations. So maybe you can try to add a flag “-fix_disulf” to see whether this will help with your design. Check the manual for details of using this flag.
And, what do you mean CS-Rosetta?
> Dear developers,
> I am running CS-Rosetta on a protein with a known disulfide pattern and a large proportion of flexible structure. The first models look promising in terms of the recognition of some secondary structure elements (there is a X-Ray structure for the protein), however the overall tertiary structure perdiction is rather poor, which is due to the large flexible links between the secondary structure elements I believe. If I was able to use the known disulfide pattern as restraints the prediction would probably improve significantly. In my current setup the cysteines could not form bonds at all because they are too far from each other. Is there some way to change this? I found there’s a disulfide filter mentioned on the hoempage but could not figure out how I can possibly change its setting.
> Many thanks and kind regards
> Baerbel Blaum