You are here

'-norepack_disulf' flag NOT work in Ligand-protein docking ?

7 posts / 0 new
Last post
'-norepack_disulf' flag NOT work in Ligand-protein docking ?
#1

Dear All :

I met some problems when I try to dock a small molecule to a protein using the command "ligand_dock.linuxgccrelease"

What I want:

repack most of the sidechain of protein, but there are several disulfide bonds in the proteins, and I try to use the Flags '-fix_disulf' to
find the disulfide bonds and use '-norepack_disulf' to freeze the CYS involved in the disulfide bonds. the Flags file includes flowing lines :

-extra_res_fa BSP.params
-nstruct 2
-fix_disulf
-norepack_disulf
-out:file:renumber_pdb -no_optH
-ex1 -ex1aro
-randomize2
-uniform_trans 5
-minimize_ligand
-harmonic_torsions 10
-run:constant_seed -nodelay
-start_from 23.60 35.43 21.41
-start_from 17.13 38.18 14.53
-start_from 23.18 31.04 26.60
-s insulin.pdb

the command I use is "/home/mldaet/rosetta_old/rosetta3_source/bin/ligand_dock.linuxgccrelease -database /home/mldaet/rosetta_old/rosetta3_database/ $flags"

problems :

the docking results show that the disulfide bonds are not freezed as I want, it also repacked as other side chains. I also tried the flags "-norepack1" as a control
and the results show that that flag seems not working, all the side chain of proteins are repacked. Does this mean the ligand-protein docking in rosetta will consider
all the side chains' repacking by default ? Any suggestions are deeply appreciated !!!

Best Regards to all.
Da Lintai
HKUST

Post Situation: 
Tue, 2011-01-11 22:21
darlintai

You could also try passing a resfile that says NATRO at the disulfide positions. Many of the flags that you mentioned are not universally respected flags; they are only functional for a few protocols.

I've contacted one of the ligand docking authors to get more expert input.

Wed, 2011-01-12 07:00
smlewis

Thanks for your reply but unfortunately it seems not work either, I add the flag '-packing:resfile resfile', and input a file named "resfile" including the lines :

6 NATRO
7 NATRO
11 NATRO
28 NATRO
20 NATRO
40 NATRO

the protein have three disulfide bonds.

Further suggestions are appreciated.
Best Regards.
Da Lintai
HKUST

Wed, 2011-01-12 07:46
darlintai

I think the NATRO command should work. I think that you need to specify the chain ID though, like this:
6 A NATRO

would keep the native residue in residue 6, chain A. When you use the NATRO command the repacking system completely ignores it. A side effect is that the residues you specify as NATRO will not be included in energy calculations during packing. This is kind of a bug, but in practice this does not usually have a major effect on the system.

What smlewis originally said about the fix_disulf option is correct, it is not implemented in the ligand docking application and will therefore have no effect at all.

Wed, 2011-01-12 14:34
delucasl

Er, in light of Sam's reply, if you used this resfile and it didn't crash, it probably never read it in. It ought to crash since you forgot the chain letter.

Wed, 2011-01-12 14:42
smlewis

Thanks for the hints, But it still does not work even I add the chain ID, the Flags I using are :

-extra_res_fa BSP.params
-nstruct 10
-out:file:renumber_pdb -no_optH
-randomize2
-uniform_trans 5
-minimize_ligand
-harmonic_torsions 10
-packing:resfile resfile.log
-run:constant_seed -nodelay
-start_from 23.60 35.43 21.41
-start_from 17.13 38.18 14.53
-start_from 23.18 31.04 26.60
-s insulin.pdb

in the resfile.log file :

6 A NATRO
7 A NATRO
11 A NATRO
28 B NATRO
20 A NATRO
40 B NATRO

because there are two disulfide bonds are inter-chain, I also tried two of them:

6 A NATRO
11 A NATRO

the results show that the disulfide bonds are still not totally freezed.

Best.

Wed, 2011-01-12 20:44
darlintai

It seems likely to me that it's just ignoring your resfile. Probably whoever wrote it never thought of having disulfides nearby.

You could try just resetting those residues back to their appropriate conformations after the ligand docking application is finished, then checking their score (score.cc) to make sure they don't clash too bad.

You could try going through the code looking for spots where packing occurs, and we can stick in code to make it respect the disulfides (I'll give more details if you want to try this)

You can try waiting for the rebooted version in 3.2; I think it's been reworked with the RosettaScripts such that it will be more flexible.

Tue, 2011-01-18 07:54
smlewis