How many steps and replicas is sufficient for my peptide?
The longer the peptide is, the more difficult is the docking problem. We always suggest running more replicas with more steps if resource allow. A guideline should be running 100-300 replicas with a maximum Monte Carlo steps of 1-3 million per amino acid in the peptide (10-30 million steps for a 10-mer peptide).
Do I start from a helix, an extended structure, or a mixture of both?
If you have prior knowledge of your peptide’s conformation. We suggest to start from that conformation. Starting from one conformation does not prevent the sampling of other conformation with different secondary structure. If nothing is known about the peptide, we suggest to start from a combination of 80% extended conformation and 20% helical conformation.
Why there are clashes in the output structure?
ADCP uses a discrete rotamer library to construct the sidechain and a coarse-grained scoring function for the internal energy. Thus, the output structure might have some minor clashes as well as nonoptimal bond angle and bond length. We suggest running a minimization with any MM force fields to resolve these minor issues. We will provide a all-in-one solution in the future.
Where are the caps at the peptide termini?
We discovered that there are different types of both peptide termini. We also found adding the proper caps does not conclusively help the docking. Thus, we decide to keep it simple by removing the caps while docking. Adding back the caps to the output structure can be done using software like VMD or PyMOL.