You are here

side chain repacking error

5 posts / 0 new
Last post
side chain repacking error
#1

Hello!

- I tried fix_bb script and used a resfile > 1 A NATAA, 2 A NATAA and so on as per sample resfile but it shows an error "core.pack.task.ResfileReader: On line 2, the chain identifier 'NATAA' must be just a single character in [_A-Za-z] (note the chain identifier is case sensitive)." I could not understand it.

-Moreover, is it necesaary to make a resfile having information for indiviudal amino acids. Isn't there any option or flag which can fix the amino acids positions and only repack teh side chains (e.g. at all positions I need same amino acids and don't want sequence to change but prefer to have repacking of their side chains).

- second, I am getting a lot of confusion and trouble between using minimizer and relax. Minimizer is a subpart of relax alongwith 'packer' so there must not be any confrontation between both. For saying, in the past I was using relax on a protein with ligand and metal, in which my ligand was totally debonded (totally messed up) and thrown out of the pocket. Then, I shifted to minmize_with_cst and it worked. This time again, when I am trying the 'relax script' on some new protein using following flag   "relax.static.linuxgccrelease -s 1xkk.pdb -nstruct 1 -relax:bb_move false -use_input_sc -relax:default_repeats 1 -extra_res_fa FMM.params -database /home/gnss/singhma/rosetta/software/rosetta_bin_linux_2017.26.59567_bundle/main/database/ -out:level 400 -constraint:cst_file cstfile &" again getting a debonded ligand which looks absurd. However, it works fine if I use minimizer (specifically minimize_with_cst).

Here the ligand used is multiple copies of the crystal ligand with proper bond order and valencies (prepared) as I dont want my ligand to vary too much from its original crystal conformation. 

Could you suggest me any solution so that I could use relax protocol of rosetta for my proteins (having ligand and metals). Indeed, I am interested in repacking of protein side chains before any future applications and most importantly I don't want my ligand to enjoy outside the pocket which is commonly happening when i use relax. 

 

Category: 
Post Situation: 
Thu, 2017-12-07 04:49
malkeet.singh

check this output file for previous message (for ligand)

File attachments: 
Thu, 2017-12-07 05:29
malkeet.singh

resfile: it's whitespace delimited; the first column is the residue number, the second is the chain, and the rest are commands.  The error says that "NATAA" is appearing in your second column.  So apparently the chain ID is missing in at least one line.

using the resfile:

1) The resfile controls packing.  Packing NEVER moves the backbone, the whole resfile documentation should be read with that in mind - we are only discussing the sidechains' movement.  NATRO is "don't move the sidechain", NATAA is "move the sidechain but keep the sequence", and many of the other commands let you change the sequence.

2) The top of the resfile can specify a default, which is applied to all positions not otherwise specified.  So for a 100 residue protein, you can specify separate behavior at positions 4, 65, and 72; and the default will apply to the rest.

Your ligand getting thrown out:

Your ligand is probably being treated by Rosetta as a free chemical species.  If it's getting ejected from the pocket it's because Rosetta thinks it's got clashes.  The simplest solution is to use constraints to keep it in place.  Relax offers automatic constraint modes (see its documentation).  Setting NATRO in a resfile for the residues lining the pocket will have a similar effect if you are doing just repacking.

If you mean for your ligand to be chemically bound to a sidechain, you need to set up a separate set of residue parameters for that - params files.  You'll see lots of questions about this on these forums.

Thu, 2017-12-07 13:41
smlewis

Hello!

 

1. NATAA is "move the sidechain but keep the sequence" >> what do you mean by keep the sequence here?

2.  The top of the resfile can specify a default, which is applied to all positions not otherwise specified.  So for a 100 residue protein, you can specify separate behavior at positions 4, 65, and 72; and the default will apply to the rest >> I tried by putting default in resfile but it didn't work. following is the error.

core.pack.task.ResfileReader: On line 1 command 'DEFAULT' is not recognized.
ERROR:: Exit from: src/core/pack/task/ResfileReader.cc line: 1518
 

3. Setting NATRO in a resfile for the residues lining the pocket will have a similar effect if you are doing just repacking. >> Are you sure? Won't amino acids lining binding pocket re-pack in 'packer' application as contrarily NATRO will keep them intact/immobile ? 

4. My ultimate aim is to check the effect of Rosetta side chain repacking (binding site) on docking results so Q3 is imperative. 

 

Thank!
Malkeet

Wed, 2017-12-13 07:30
malkeet.singh

"what do you mean by keep the sequence here?"

From the packer's point of view, all sidechain rearrangements are the same operation. It will test lysine-rotamer-1 to lysine-rotamer-10, or it will test lysine-rotamer-1 to arginine-rotamer-32.  It doesn't care if the sequence changes or not.  NATAA says "repack only to rotamers of the same type of residue" - don't let the protein sequence change.

DEFAULT:

There is no "DEFAULT" command; whatever is present before the start token is the default, as the https://www.rosettacommons.org/docs/latest/rosetta_basics/file_types/resfiles#resfile-syntax-and-semantics_header documentation explains.

 

NATRO:

If you are only repacking - like using the fixbb app - then there are already no degrees of freedom for the ligand to move; setting the pocket around it to NATRO will also ensure that those residues do not move.  If your goal is "do not move the ligand or pocket residues under any circumstances" this should behave that way.  It sounds like you want to repack so I guess this is irrelevant.

Wed, 2017-12-13 09:54
smlewis