Hi, novice Rosetta user here (so apologies in advance for simple / silly questions).
I have small peptide ligands that I am using in standard grafting / design. This has been relatively straightforward. However, one monkey wrench I'd like to throw into the works would be for the design scripts to use a noncanonical amino acid instead of the corresponding canonical one. For example, instead of tyrosine, design interfaces assuming all tyrosines are phosphorylated, or alternatively, impose acetylation or methylation on certain residues.
I found a forum post here (https://www.rosettacommons.org/node/3556) that mentions writing a "patch" file (in this person's case, lysine covalently modified with retinal). Such an approach looks like it might work for me. I just don't know how to either write the patch file, or incorporate it into the design scripts.
Any help or examples, whether on the above "patch" approach or a more appropriate approach, would be much appreciated. Thank you!
Rosetta already has patches for some of the common residue modifications. These include phosphorylation (STY), sulfation (Y), methylation (K), acetylation (KC) and hydroxylation (P). These should be loaded automatically if your input PDB has the appropriate residues. Basically, your input PDB just needs to have at least one heavy atom from the modification present. (Matched by name -see the files in main/database/chemical/residue_type_sets/fa_standard/patches/ for indications on the Rosetta naming convention.) As these are sidechain atoms, their positions should be replaced by ideal conformations when these residues get repacked. (Omit the "use_input_sc" setting of the packer to ensure that you don't accidentally keep that residue.)
There's not really a good way of controlling adding or designing to those residues without editing the PDB, however. For example, I don't know of a resfile command which will specify that you want to design to a particular patched variant.