# protein protein docking pipeline

5 posts / 0 new
protein protein docking pipeline
#1

After going through the tutorials of protein protein docking, I totally got in a mess.
The flag file and XML file, low resolution and hign resolution docking, clustering seem to mix together.
e.g.
1. $docking_protocol.linuxgccrelease -database$ROSETTA_DATABASE @flag > log.txt
The flag file as follows:
-parser:protocol low_res_docking.xml
-s input.pdb
-docking
-dock_pert 8 5
-spin 1
-randomize2
-docking_centroid_outer_cycles 50
-docking_centroid_inner_cycles 500
-docking:dock_mcm_trans_magnitude .1
-docking:dock_mcm_rot_magnitude 1
-nstruct 10
-linmem_ig 10
-ex1
-ex2
-ex1aro
-overwrite
-packing:repack_only
-out:pdb

The low_res_docking.xml contents as follows:

2.It seems that the low resolution docking process encompasses high resolution docking as in the XML file.
So then I don't know how to make a high resolution flag file and XML file.Could you please list a general pipeline of protein protein docking? And is there any reference flag and XML file?

I am also not sure about the important parameters that I should set according to my protein. The manual seems not so much comprehensive...

Post Situation:
Thu, 2012-07-19 06:25
dzhao

The low_res_docking.xml file could't be shown.
Attached is its content.

Thu, 2012-07-19 06:35
dzhao

OK, first thing: there are two ways to do docking, and (as you suspect) you are accidentally trying both.

One option is docking_protocol. This exclusively uses command line parameters to control how docking occurs.

The other option is rosetta_scripts, also called the parser. This uses an XML file to control some things, and uses command line flags to control other things. The XML version is much more customizable in the long run, but slightly harder to get started.

1. $docking_protocol.linuxgccrelease -database$ROSETTA_DATABASE @flag > log.txt
The flag file as follows:

#-parser:protocol low_res_docking.xml #removed - does not work with docking_protocol
-s input.pdb
-docking

#use only one of these three flag sets at a time
#-dock_pert 8 5
#-spin 1
-randomize2
-randomize1

#removed - leave at default
#-docking_centroid_outer_cycles 50
#-docking_centroid_inner_cycles 500
#-docking:dock_mcm_trans_magnitude .1
#-docking:dock_mcm_rot_magnitude 1

#nstruct 10 is ok for a quick test, but you need at LEAST 10,000 for a real run
-nstruct 10

-linmem_ig 10

-ex1
-ex2
-ex1aro

-overwrite #be careful with overwrite, it overwrites old results
-packing:repack_only
-out:pdb

Thu, 2012-07-19 07:52
smlewis

Thanks for your reply! I'm still not quite understand the whole docking process.
1.Is the docking protocol you list the low resolution one? Then after the protocol above, should I turn to a hign resolution protocol? Then is there another high resolution flag? Need I insert a clustering process?
2.Rosetta needs to repeat docking low and docking high to get the whole minimum energy state. How does the repeating process relized? Where is the parameter of repeating times set?

Fri, 2012-07-20 00:04
dzhao

"1.Is the docking protocol you list the low resolution one? "

yes

"Then after the protocol above, should I turn to a hign resolution protocol? Then is there another high resolution flag?"

It does both together from that command line.

"Need I insert a clustering process?"

I don't.

"2.Rosetta needs to repeat docking low and docking high to get the whole minimum energy state. How does the repeating process relized? Where is the parameter of repeating times set?"

The docking executable has a pre-decided optimum number of cycles, based on the developers' experience. I guess if you want to change the number of cycles, you can, but I've never tried. These flags you had earlier are part of it:
-docking_centroid_outer_cycles 50
-docking_centroid_inner_cycles 500
There are probably analogous docking_fullatom_outer_cycles flags or something similar.

Sat, 2012-07-21 11:57
smlewis