Dear Rosetta developers,
I am interested in the de novol design of a peptide with helical conformation which could bind a specific protein partner with enhanced affinity. However, when consulting the user guide, i lost my way about the specific method. So could you be so kindly to tell me which method or module in Rosetta can do such kind of case , or provide me a step-by-step guidance.
What i intended to do was listed in the following:
1. Construction of initial models with helical backbone conformations. I hope these models could contact its partners with one helical surface comprised of hydrophobic residues.
2. For each of the models, iterating between sequence optimization and backbone/sidechain refinement for several rounds.
3. Scoring the final models and ranking.
I guess several modules in Resetta may employed simultaneously for above-mentioned steps. Any suggestions will be greatly appreciated.
In general, for a desired protocol this precise and complex, your best bet is to write your own in RosettaScripts.
For the first problem, building a new helix-protein interface, you can probably use FlexPepDock. I don't think there's any released code in Rosetta that will build a helix in, so maybe you can do that part by just stealing a helix you like from some other structure.
For the second part, if FlexPepDock has a design mode, use that. If not, do cycles of FlexPepDock + fixbb, or relax+fixbb.
For the final part, most Rosetta modules print scores with structure outputs, so unless you need an exotic score, it will already have been done by part 2.
Thanks for your great help. It seems that there's a long way to go.
However, in the Supporting information of the JACS paper (J. Am. Chem. Soc. 2011, 133, 4190–4192) , the author used fragment-based backbone assembly to build an alpha helix. And the author mentioned that "Loop modeling in centroid mode with Rosetta revision 29304 was used to perform fragment assembly and build the peptide backbone conformations". So i am curious about whether the loop modeling method in Rosetta 3.4 could do so, although i didn't find relevant information in the user guide.
For the second part, i guess relax+fixbb or FlexPepDock+fixbb may just optimize the backbone and sidechain conformation after the whole sequence design phase, does these combination can optimize the structure on-the-fly (such as optimize the structure once fixbb mutant a residue )?
Deanne did that work in Rosetta++ (2.x, probably 2.1 or earlier, but that's a guess). I don't think she did it in Rosetta3 at all, and I know it wasn't in 3.4. The problem is not that the fragment-based assembly can't be done, it's just that it's not trivially accessible from command line, and I don't know if her protocols are widely available. I've asked the authors (they're from my lab) for more input.
Separate runs of relax/flexpepdock + fixbb will not perform on the fly optimization. If you want to tightly mix cycles of backbone rearrangement, your best bet is probably to write a RosettaScript with your precise needs. Nobody has nailed the "best" way to do this problem, so none of our methods are generally applicable, and none of the existing executables I can think of will readily extend themselves to this problem. (Actually, I just thought of one that might, let me see if it is published...)
Did you look in the "Peptide Designs" folder of the Protocol Captures in the 3.4 release? It should cover that JACS paper.
Doug reminds me that there is code for flexible-backbone peptide design (for the latter step):
"Yes. DougsDockDesignMinimizeInterface. It was released in 3.3 and has a protocol capture from the special collection. It would require some work to be generalized.
smlewis, Thanks for your kindly help！And it's great that your lab contributed so much to the rosetta_design code.
I do find the protocol for Peptide_Designs, it seems that i need install Rosetta++. At the software download page, the newest version is Rosetta++2.3.1, i don't know whether it's compatible with Rosetta 3.4 which i had installed on my computer.
I don't find "DougsDockDesignMinimizeInterface" under the protocol folder, but i find a similar one under the folder of resetta_demos/doug_dock_design_min_mod2_cal_cal, do you mean this one?