Tutorial for using modified residue types in centroid-level applications.

KEYWORDS: CORE_CONCEPTS NONCANONICALS

In this tutorial, you will describe how to wrangle modified residue types through combined centroid/fullatom protocols in Rosetta.

Background

Many Rosetta protocols use a centroid phase followed by a fullatom phase. However, the centroid residue set is compatible with the canonical 20 amino acids, and very little else. For example, ligands and post-translational modifications are incompatible with centroid mode; Rosetta generally crashes when trying to switch these residues to centroid (or silently drops the post-translational modification).
You have been provided with a PDB file containing a two phosphoresidues as input (one p-SER and one p-TYR). This tutorial will show you how run the docking_protocol application using this PDB model as input without Rosetta exiting with an error.

Clean up the PDB file

This is an NMR model. Rosetta will read this as a giant complex of superimposed structures, which is bad. You need to find the line labeled "ENDMDL" and delete all lines below it.
```
gunzip -c starting_files/2lax.pdb.gz > rosetta_inputs/2lax_edited.pdb
<your favourite text editor> rosetta_inputs/2lax_edited.pdb
```
Rosetta wants the phosphorylated residues to be named the same as their canonical countertypes. Open the pdb file rosetta_inputs/2lax_edited.pdb for editing. For residue 202 and residue 206, rename the three-letter code "TPO" to "TYR" and "SEP" to "SER". Rosetta will identify these as phosphorylated based on their atom names.
- note: If at this point you were to run any protocol that utilizes centroid mode, you would see this error when Rosetta crashed:
```
can not find a residue type that matches the residue TYR_p:phosphorylatedat position 38

ERROR: core::util::switch_to_residue_type_set fails
```

Create centroid residue parameter patch files

For modified residues (e.g. phosphorylation or acetylation), Rosetta uses "patch" files to modify the pre-existing residue parameter files. Unfortunately, a centroid-level patch file for phosphorylated residues does not exist. You need to create two, one for P-TYR and one for P-SER. 1. Go to the database directory <my_rosetta_directory>/rosetta_database/chemical/residue_type_sets/centroid/patches/ and copy the file tyr_sulfated.txt to tyr_phosphorylated.txt. Open that file for editing. Change these lines:

NAME sulfated
TYPES SULFATION

to this:

NAME phosphorylated
TYPES PHOSPHORYLATION

NOT VARIANT_TYPE PHOSPHORYLATION

Now copy this new file tyr_phosphorylated.txt to ser_phosphorylated.txt. Open ser_phosphorylated.txt for editing. Change this line:

AA TYR

to this:

AA SER

Now, we have patch files for our centroid-level phosphorylated residues. All we have to do now is point Rosetta to these files. Open the file <my_rosetta_directory>/rosetta_database/chemical/residue_type_sets/centroid/patches.txt. Add these two lines to the bottom of the file:

patches/tyr_phosphorylated.txt
patches/ser_phosphorylated.txt

Now run the docking_protocol application like this: (where $ROSETTA3=path-to-Rosetta/main/source)

$> $ROSETTA3/bin/docking_protocol.default.linuxgccrelease -s rosetta_inputs/2lax_edited.pdb

This application should now work correctly, converting the input pdb coordinates to a centroid model, performing rigid-body docking, then converting back into a full-atom model ( which should now correctly convert the phosphorylated residue types ) before performing a final docking rotamer packing optimization (see the docking_protocol documentation for more information).